Welcome to TSMC's second quarter 2014 earnings conference and conference call. This is Elizabeth Sun, TSMC's Director of Corporate Communications and your host for today. Today's event is webcast live via tsmc.com. If you are joining us through the conference call, your dial-in lines are in listen-only mode. As this conference is being viewed by investors around the world, we will conduct this event in English only. The format for today's event will be as follows. First, TSMC Senior Vice President and CFO, Ms. Lora Ho, will summarize our operations in the second quarter, followed by our guidance for the third quarter. Afterwards, TSMC's Chairman, Dr. Morris Chang, will provide a couple of key messages. We will begin the Q&A session where TSMC's two Co-CEOs, Dr. Mark Liu and Dr. C.C. Wei, will join Chairman Chang and Ms. Ho in responding to your questions.
For those participants on the call, if you do not yet have a copy of the press release, you may download it from TSMC's website at tsmc.com. Please also download the summary slides in relation to today's earnings conference presentation. As usual, I would like to remind everybody that today's 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. Please refer to the safe harbor notice that appears on our press release. Now, I would like to turn the podium to TSMC CFO, Ms. Lora Ho.
Thank you, Elizabeth. Good afternoon, everyone. Thank you for joining us today. I will start my presentation with financial highlights for the second quarter, followed by the guidance for the third quarter. Second quarter 2014 is a record quarter for TSMC, both in revenue and in net income. In this quarter, we saw strong demand for our wafers across all major segments. Our revenue increased 23.5% sequentially and 17% year-over-year to reach TWD 183 billion. On the profitability side, gross margin was 49.8%, 2.3 percentage points higher than that in the first quarter. The higher margin was mainly driven by higher capacity utilization, partially offset by inventory valuation adjustments and a small margin dilutive effect from 20 nanometer ramp. Operating margin was 38.6%, up 3.2 percentage points from the first quarter, reflecting an improved operating efficiency for the company.
During the second quarter, we recorded a TWD 2 billion disposal gain from the sale of Vanguard International Semiconductor Corporation's shares. We also saw a big jump in the corporate tax rate from 10.4% in the first quarter to 19.8% in the second quarter. The 9.4 percentage point increase was attributed to the accrual of retained earning tax, which is incurred in the second quarter every year. We estimate our full year tax rate for 2014 will be about 13%. Overall, the second quarter EPS was TWD 2.30, and ROE was 26.9%. Let's take a look at the revenue breakdown by application. Our revenue from all major segments increased sequentially in the second quarter. Compared to the first quarter, communication increased 19%, computer increased 14%, consumer increased 31%, and industrial-related revenue increased 25% during the second quarter. By technology, our 28 nanometer grew more than 30% from the previous quarter.
Its contribution to our total wafer revenue increased from 34% in the first quarter to 37% in the second quarter. Combining the two advanced technology nodes, 28 nanometer and 45 nanometer, we derived 56% of total revenue in the second quarter. Let's move to the balance sheet. Cash and marketable securities reached TWD 315 billion at the end of the second quarter. Our cash increased TWD 23 billion during the quarter, while our marketable securities increased TWD 58 billion as we reclassify certain long-term investments to short-term marketable securities. Current liabilities increased by TWD 72 billion as we accrued TWD 78 billion for the cash dividends in the second quarter. Accounts receivable turnover days increased five days to 40 days. As our revenue increased at a much faster pace than average accounts receivable in the second quarter, days of inventory decreased one day to 51 days.
Let me make a few comments on cash flow and CapEx. During the second quarter, we generated TWD 82 billion cash from operations, invested TWD 73 billion in capital expenditure, and borrowed TWD 10 billion short-term loans for hedging purposes. At the end of the second quarter, our cash balance increased TWD 23 billion to reach TWD 265 billion. Free cash flow for the second quarter was an inflow of TWD 80 billion. In US dollars, our second quarter CapEx was $2.4 billion. That made our CapEx total $6.2 billion for the first half of the year. Our full-year CapEx budget remains in the range between $9.5 billion and $10 billion. Regarding our capacity, we now expect to increase our capacity by 11% from last year. This is a small increase from the 10% growth as we forecasted earlier, and it is mainly derived from our productivity improvements.
Total annual capacity is expected to reach 8.1 million 12-inch equivalent wafers this year. I have finished my report on the financial part. Let me turn to the third quarter outlook. Based on our market outlook and the business exchange rate and forecast exchange rate of 29.81, we expect our third quarter revenue to be between TWD 206 billion and TWD 209 billion, representing between 12.6% and 14.2% sequential growth. Gross margin is expected to be between 48.5% and 50.5%, and operating margin is expected to be between 38.5% and 40.5%. This concludes my remarks. Let me turn the podium to our chairman, Dr. Morris Chang.
Good afternoon, ladies and gentlemen. I want to make a few comments on the supply chain inventory situation and the second half demand, and then a few words on 20 nanometer and 16 nanometer progress, and then a few words on 10 nanometer progress. First, on supply chain inventory and second half demand. Our second quarter was record-breaking in revenue, operating income, and net income. We have noted in past investor conferences that supply chain inventory dipped to a low point at the end of last year. The past six months have seen a rebuilding of inventory, which partially explains the strength of demand in that period. We estimate that the supply chain inventory has returned to seasonal level at mid-year. From a very low point at the end of last year to seasonal level at mid-year.
We further estimate that it will actually continue to build through the third quarter. DOI will be two days above seasonal at the end of the third quarter. Then we estimate that DOI will decrease and will be at two days below seasonal at the end of the year. The last six months have seen a rebuilding of inventory, and at mid-year, it's about seasonal level. It will continue to build some modestly to two days above seasonal at the end of the third quarter, and then it will dip to two days below seasonal at the end of fourth quarter. One might characterize the next six months as a period of cautious inventory management by the supply chain. However, the widespread adoption of 4G LTE tends to increase the demand of our 28 and 20 technologies.
Overall, we're guiding third quarter to be another record-breaking quarter, as Lora just said. It will be another record-breaking quarter in revenue, operating income, and net income. Also, at this point, because of our strength in 20SoC and our strength in 28, 4Q appears to be a sequential growth quarter also. A few words on 20 nanometer and 16 nanometer progress. In the last two and a half to three years, 28 nanometer technology has driven our growth. In the next three years, 20 and 16 nanometer technologies are going to drive our growth. 28 in the last two and a half to three, 20 and 16 in the next three. After two years of meticulous preparation, we began volume shipments of our 20 nanometer wafers in June. The steepness of our 20 nanometer ramp sets a record.
We expect 20 nanometer to generate about 10% of our wafer revenue in the third quarter, and more than 20% of our wafer revenue in the fourth quarter. We expect the demand for 20 nanometer will remain strong and will continue to contribute more than 20% of our wafer revenue in 2015. We will reach 100% of our total wafer revenue in the fourth quarter of this year, and it will be above 20% of our total wafer revenue next year. The 16 nanometer development leverages off 20SoC learning and is moving forward smoothly. Our 16 nanometer is more than competitive, combining performance, density, and yield considerations. 16 nanometer applications cover a wide range, including baseband application processors, consumer SoCs, GPU, network processors, hard disk drive, FPGA, servers, and CPUs.
Volume production of 16 nanometer is expected to begin in late 2015, and will be fast ramped up in 2016. The ecosystem for 16 nanometer designs is current and ready. A few years ago, in order to take advantage of special market opportunities, we chose to develop 20SoC first, and then quickly follow with 16 nanometer. We chose this sequence to maximize our market share in the 20/16 nanometer generation. As the 20/16 foundry competition unfolds, we believe our decision to have been correct. Number one, in 20SoC, we believe we'll enjoy overwhelmingly large share in 2014, 2015, and onwards. Number two, in 16 nanometer, TSMC will have a smaller market share than a major competitor in 2015, but will regain leading share in 2016, 2017, and onwards.
Number three, if you look at the combined 20 and 16 technologies, TSMC will have an overwhelming leading share every year from 2014 on. Number four, in total foundry market share, after having jumped four percentage points in 2013, TSMC will again gain several percentage points in 2014. This is the total foundry market share covering all technologies. After having increased four percentage points last year, TSMC will gain another several percentage points this year. A few words about 10 nanometer. The 10 nanometer development is progressing well. The 10 nanometer speed is 25% faster than the 16 nanometer. The power consumption is 45% less than 16 nanometer, and the gate density is 2.2x that of the 16 nanometer. Power is 25% faster. Did I say speed? Power? I meant speed. Speed is 25% faster, power is 45% less, gate density 2.2x more, all compared with 16 nanometer.
We work closely with our key customers to co-optimize our 10 nanometer process and design. We expect to have customer tape-outs in the second half of 2015. Those are all my prepared comments. I believe we are ready for Q&A.
Yes. This concludes our prepared statements. Before we begin the Q&A session, I would like to remind everybody to limit your questions to two at a time to allow all participants an opportunity to ask questions. Questions will be taken both from the floor and from the call. Should you wish to raise your questions in Chinese, I will translate that to English before our management answers your question. For those of you on the call, if you would like to ask a question, please press the star then one on your telephone keypad now. Questions will be taken in the order in which they are received. If at any time you would like to remove yourself from the questioning queue, please press the pound or the hash key.
Here's one already.
All right. I think Chairman wants Randy, right?
Yeah.
Okay. Okay. Credit Suisse, Randy Abrams.
Okay. Thank you.
Randy, yeah.
I appreciate you calling me first. I'm sorry to harp on one negative with all the results quite good. I wonder if you could elaborate on your comments on 16 nanometer, where you talked about your market share next year being lower than a competitor. If you could talk about why you think that plays out, the potential impact of that, and then confidence to regain share. What's driving confidence to regain share the following year?
Will you repeat that question?
Sure. Randy's question is with respect to Chairman's comment on 2015's market share at 16 nanometers, where we said that we will be lower than a major competitor in 2015. Randy's asking why we will be lower, and what is the impact to TSMC if we have a lower market share, and what give us the confidence that we will regain the market share in following year?
Okay. We need to go back through history a little bit. 28 nanometer followed 32, that particular major competitor that I referred to chose 32 and skipped 28. Then, of course, we came to 20 and 16 for us, 14 for him. We chose to do both. Actually, we chose to do 20 first, and 16 about a year or so later. It was a pretty quick succession. This major competitor skipped 20 and went on to 16. You ask why is our market share on 16 lower than this major competitor in 2015? The answer is, we got started a little late because we chose to do 20 also, and 20 turned out to be a very, very important node, as I already spoke to just a few minutes ago. We have both 20 and 16.
How can we be sure that we'll regain 16 share in 2016? Well, the answer is that with technologies these days, you just have to engage with the customers way in advance, like about two years in advance, before you get orders. You discuss with the customers, the customers follow your technical progress, and then the customers give you tape outs, and you have to qualify the tape outs, et cetera. The whole process is about two years in advance of actual volume delivery. How can I be sure that we'll regain share in 2016? Well, this is middle of 2014, about two years ahead of 2016. I think I have a pretty good idea of how much order I'm going to get in 2016.
Should the impact to TSMC, if you could talk about how material the revenue impact of that market share, because you'll be doing 20 in high volume, to how much impact, potential to outgrow the industry or how much revenue impact from this market share loss on 16 initially?
Could you repeat that question, Ruby?
Randy, your question is, in 2015, since we will lose some market share at the 16 nanometer, so what's the impact to us in revenue if we do not have that part of the business?
Ability to outgrow the industry like you have the past few years. Yeah, either revenue impact or how you see your growth relative to industry.
Will we be growing faster or slower than the industry next year?
Well, I don't know. I think there's too much unknown still. I think I have a pretty good idea of how much revenue our competitor, foundry revenue now I'm talking about, how many foundry revenue our competitor is going to have in 2015 on the 16 nanometer. I don't think I want to publicly say what the number is. Actually, there is still a bit of uncertainty in the number. We could get some of it. We are going to be late, but we'll be there, actually, in the second half.
Randy, maybe we should remind you that Chairman said in his key message that 2015, our 20 nanometer will be accounting for more than 20% of our total revenue.
Yeah.
I think that would give us-
I think Randy understood that, and I think everybody also does so.
Okay. Can I just get a second? Yeah. Okay.
You have three questions.
Yeah, you got two chances already, Randy. I think you wait your turn now and later.
All right. Okay. Next question, I think, would be J.P. Morgan's Gokul Hariharan.
What's her name?
Gokul.
Gokul.
It's Gokul.
Okay. Gokul, did you say?
Gokul.
Okay. All right.
Thanks for taking my question. I had a couple of questions on 10 nanometer. Since you've mentioned that 10 nanometer will start this production in Q4 of next year, is it fair to assume that 16 nanometer as the leading edge-
I'm sorry. I need to interrupt. Did you say that we said before when 10 nanometer-
Chairman, you just mentioned that 10 nanometer would start this production in second half of next year?
No.
After next.
10? No. 16 will start volume production.
In 2016.
in second half of next year.
Okay.
I didn't say anything about 10. Don't misunderstand. That's a very important distinction, 10 and 16.
Yeah. Just wanted to understand what is the cadence for 16 nanometer, since we are starting a bit later, since 20 is still going to be more than 20% of revenues next year.
10.
Does 16 nanometer also ramp up to be 20-plus % of revenues in 2016, or is it going to be a shorter node and being overtaken by 10 nanometer maybe sometime in 2016 or 2017?
You will have to repeat that, Martha.
Okay. Gokul's question is with respect to the cadence. First part is with respect to the cadence between 16 nanometer and 10 nanometer, whether or not we will have a shorter cadence or a regular cadence. Whether or not 16 nanometer will account for more than 20% of the revenue for us in 2016, or 16 nanometer will become a short node.
What's the second question?
Whether 16 nanometer will account for more than 20% of our wafer revenue in 2016, or whether or not 16 nanometer will be very short.
Whether 16 nanometer Say it again, please. Yeah.
Whether 16 nanometer will be 20% of revenue by 2016.
2000?
16.
16.
Correct.
I don't think I've looked at it. Have you? 2016, I don't know what the total Well, I think the total revenue will be higher than 2015, but exactly what it will be, I don't quite know. Really, I have more problem with your question. I have more problem with the total revenue of 2016.
If I rephrase it.
16 nanometer revenue in, yeah.
If I could rephrase it, will 16 peak out lower than 20 nanometer?
No. I have a feeling that 16 will be as strong or stronger node for us than 20 nanometer.
Okay. The second question I have is on-
I don't think we have answered your first question. Your first question was 10? What was the other one?
Cadence. The cadence between 16 nanometer and 10 nanometer, whether or not it will be a shorter cadence, or-
Well, I just noticed that Intel said that they aren't about to talk about their 10 nanometer agenda, until sometime next year, right? I'm going to follow that practice. Okay? Yeah.
The second question I had is, in terms of capital intensity. I think this year and next year, I think capital intensity is already coming down. Should we expect a change when it comes to the next leg of investment, or is it still going to be the same pattern in terms of capital intensity coming down?
No.
Capital intensity is measured by CapEx divided by revenue, whether or not this number will be coming down or stays the same.
It'll be coming down.
Okay. Thank you.
Yeah.
Okay, Dan Heyler, the next one from BofA Merrill Lynch.
Thank you. Thank you, Chairman and team. I wanted to follow up a little bit on, again, back to the 16 question. It feels a little bit like, I remember the CPU days where you'd had somebody who'd fallen behind on one node and then try to leapfrog. I want to revisit this question of the probability of success in skipping a node. TSMC has been, over the years, relatively careful at making sure that you're bringing up your technology and learning something at a time. We have double patterning, obviously, at 20 nanometer. And now we have FinFET at 16 and 14, in addition to two shrinks. Have we ever seen a company successfully have basically three major changes at once, leapfrogging and being able to do that? I'm just wondering, you have consciously chosen 20 for multiple reasons. I think-
I'm sorry, what three changes are you speaking of?
We have FinFET, double patterning, and also you have a couple of node shrinks. Your competitors-
You have double patterning.
Scaling.
Shrink, scaling.
Scaling.
What is your confidence that in fact, that can be done successfully by your competitor?
By our competitor?
Which is.
I'm not going to comment on how successful they will be. No.
You've just done that.
Well, I don't think it's going to be easy. As far as we are concerned, you mentioned three hurdles. To us, I think the double patterning will not be a hurdle because we have done that in 20 already. 16 will not be But your question was not about us. It was about our competitors, right?
Yes.
I don't think I really want to comment on that, except to say that it's not going to be easy.
Yeah. Well, I mean, yeah, I think you are commenting already by saying you're going to lose share. I just wanted to get a feel, maybe flesh this out. I think it's important from an investor standpoint to understand the complexity in doing this. I think now you're pretty much down the road in 20 nanometer. I think CC and Mark have had a good look at 20. You've looked at the challenges in bringing up double patterning. How critical is 20 as a learning node?
How critical is 20 as a learning aid? Yeah.
I think-
It's not working yet.
There you go.
Dan, you had a keen observation. I think indeed, I haven't seen the industry skip nodes to be that successful, if not still catching up. We never underestimate our competitors, so I wouldn't want to comment. Every change of each generation is very difficult. Therefore, you see the skills, technology, this definition, they change one thing at a time, as you know. That's most of it. For the people trying to catch up, that's what they do. There are a lot of historical data we can see.
Thank you. Second question. R&D have been impressed by your operating efficiencies, Lora. With this 16 acceleration, 10 looks pretty aggressive. Should we be modeling R&D pickup, growing it at a faster rate than sales? Or will it maintain a percentage of sales over the next two to three years?
Let me answer. If you're asking could our R&D will build up even faster than revenue, is that right? I think that it's possible. In fact, that's really our first priority. I think we want to strengthen our technical side even more. It's not going to be very big. Right now, actually, we modeled R&D at 8% of revenue at the beginning of the year. The revenue has turned out to be bigger than we thought it was going to be. Right now, R&D is only 7.3% of revenue. In my mind, however, the model is still 8. All right?
Thank you.
Next question will come from the floor. It will be from Citi's, Roland Shu.
Hi. Good afternoon, Chairman. I think I just put a still a follow-up question on 16 nanometer. I believe the Chairman comment in AGM that TSMC 16 nanometer is ahead of competitor's 14 nanometer in three regard, first is technology and productivity and also trust of customer. I think for the productivity and trust of customer, actually, we have no doubt about that. Can Chairman quantify that how big the lead it is for your 16 nanometer FinFET to competitor's 14 nanometer FinFET? Thank you.
Actually, I intended to repeat what I said in the annual general meeting just now. Really, if you combine the three, speed, density, and power, we feel that we are more than competitive. Now, you asked me how big a lead we have, et cetera. Well, I think that will be a very difficult question to answer because our competitors on the 16 are the two companies that I, in the past, referred to as the two big gorillas of the industry. Each has its strength. The two are not the same at all. Each has its strength, each has its weaknesses. A competitor may be technologically strong, but if he turns out to be a competitor of his potential customers, then that's a weakness.
If a competitor is already used to being both a competitor and a supplier, but he has got technology weaknesses, well, that's a problem, too, for him. Well, I don't think I can really answer your question in a very simple way. I can only answer it by just repeating what I said earlier. Combining performance, density, and speed. Speed, density, and power, we believe we are more than competitive. In terms of the ability to get business, we believe we are more than competitive. Okay?
Thank you, Chairman. My second question is for Elizabeth for the 3Q. The revenue guidance now is up 12%-14%, but I think now the 20 nanometer itself probably contribute more than 10%. That means then some 20 nanometer revenue growth in 3Q probably will be below seasonality.
Is this any reason behind this, or is this just because of the capacity constraint for the non 20 nanometer revenue? Thank you.
Roland's question is with respect to third quarter. Our guidance of the third quarter's sequential growth rate is 12.6%-14.2%, our 20 nanometer growth rate is going to be a lot faster, which means the non 20 nanometer geometry growth rate will be slower. What was the reason behind the slower growth rate in other nodes? Is it because of capacity constraint or some other reasons?
Randy. I'm sorry.
Roland.
Roland. Chairman was commenting on the second half. We'll go through an inventory depletion cycle again. That actually has some impact on the third quarter and fourth quarter. It's not the capacity constraint issue, it's the overall inventory level going into the third and fourth quarter.
Okay. Do you see any specific segment actually have the inventory correction issue into the second half? Thank you.
I don't want to comment. Actually, I don't have information on DUI on specific segments. Based on my guidance I just gave you earlier, we are seeing very strong growth in communication in the third quarter. However, on the computer and consumer will be a decline segment. Industrial and related product will have a modest growth.
Thank you.
Okay, we will now take our next question from the call. Operator, please proceed with the third caller on the line.
Thank you. Your first question comes from Steven Pelayo from HSBC. Please ask the question.
Yes. I apologize if this was asked, the line dropped there. Your 20 nanometer ramp is very impressive, very steep. I guess if I think about your revenues excluding 20 nanometer, going to 10% of revenues and more than 20% of revenues in the fourth quarter, it kind of suggests rather anemic growth for the industry in the third quarter and a pretty sizable down in the fourth quarter. Is that what you think the industry's doing, or do you think that maybe 28 nanometer competition and pricing pressure is starting to have an impact there? It just seems odd when I think about your revenues excluding 20 nanometer, because that really is driving so much of your growth in the second half of the year.
Steven's question is very similar to Roland's question earlier, in the sense that if we take out the 20 nanometer, then our third and fourth quarter growth doesn't look so impressive. Is that because of competition heating up at 28 nanometer, or is it because of some other reason?
I guess, if we take out I think the rest of it, if you take out 20 nanometer, I'm just calculating in my head right now. If you take out 20 nanometers, will we have growth in the third and fourth quarter? I think we still do.
We'll have growth in third quarter.
We'll have growth in fourth quarter, will we have growth in the fourth quarter? I haven't said anything about-
Fourth quarter
about fourth quarter, total fourth quarter yet. I'm just mentally calculating what it looks like if we take out 20 in the fourth quarter. The answer of my mental calculation is that we'll still have little growth. Yeah. I don't know how little is little. It may not be too little. Yeah, we still have growth, but I think just as Lora has said, and I also said it in my opening remarks, that the inventory will be cautiously managed. It almost always is toward the end of the year. Last year, in fact, it went overboard. The caution went overboard, and as a result, we had a pretty bad fourth quarter, and a lot of people had a pretty bad fourth quarter, only to recover it in the first half of this year.
We are predicting, as I said earlier, that our inventory will be very cautiously managed. Of course, we'll tell you again three months from now what it looks like at that point. Just based on the numbers that we have already given out, I would say that, yes, even if you take out 20, we'll still have growth. We'll certainly have growth in the third, and we'll still have growth in the fourth.
Okay.
That is already taking into account the cautious inventory management that will happen in the second half of this year.
Okay, great. Maybe just one quick question for Lora. I noticed in the management report, you no longer were breaking out depreciation within cost of goods sold, and I think you've guided for total depreciation to grow, I think it was 30% or 35% this year. Could you just remind us what it was within cost of goods sold in the second quarter and your outlook for the third quarter?
Okay. We have a chart in the later page, the last page of management report that shows the depreciation, including the COGS. What I can tell you, the overall depreciation for the first quarter this year is TWD 41.6 billion, and the second quarter will be TWD 45.9 billion and will go up very steeply in the third and fourth quarter. For the whole year, we should expect the year-over-year depreciation will be 33% growth, which is lower than what I have told you in the past.
Okay, slightly lower. Just so we can try to quantify, where does it step up to in the third quarter?
You mean from second quarter to third quarter, how much is the depreciation increase?
Yeah. What's the depreciation going back?
Third quarter will be TWD 59.5 billion depreciation.
59.5. Okay. Great. Thank you very much.
The next question will come back to the floor, and it will be coming from Barclays, Andrew Lu.
Hello, Chang. I have a question for you.
Okay.
Regarding the market share on the 16 nanometer you mentioned earlier, next year, we might have a lower share compared to competitor. Can we use something to describe it? Next year, do you see the total addressable market for this thing that is large, like this year, like 20, or much smaller? As either this one or the second one, we might lose shares because our 20 nanometer customers switch to competitor using 14 nanometer, and that's why we're losing share for not using our 16. These two answers will have a dramatic difference for the company's earnings prospect.
Company's earnings?
The first one is, for example, this year we have a 50,000, 20 nanometer capacity by the end of this year.
I'm trying to follow your question, let's first map a little bit. Did you get the questions?
Yeah.
The first question I think I got. The first question is a two choice. Yeah. Would you repeat it?
Two choice.
Two choice.
Andrew, your question is you want to know the overall size of the 16/14 nanometer business, overall size of the business in 2015, whether or not that is bigger or smaller than the 20 nanometer size of 2014.
That's correct.
That's your first question.
Is that our shares losing because the shares much smaller or not?
Okay. I think. Did you talk about 2015?
Yes.
Put it this way, our 20 and our 16 is much bigger than the share. It's much bigger than the 16
I think his question is, this total 16 foundry market.
Right.
Next year, is it as big as the total 20 this year? Is that your question?
Yes.
Yeah.
I think it's smaller.
Smaller? The total?
Total FinFET , including 16, 14 FinFET from older customer base.
Yes.
Compare. Okay. About comparable.
Comparable. We are losing share.
Losing what share?
Foundry shares.
Oh, that's what I said.
Yes.
Of course. We will print the share this year for either party, right? Next year, we will have a lower 16 share than a major competitor. That's what I said.
Yes. My question is, whether losing share because our 20 customer moved to a competitor for 14, or this is a brand new demand, which chose competitor's 14 nanometer first.
Yeah. As a rule, we are not going to comment on specific customers, all right?
I did not mention any customers.
Well, you said because one
It's one important customer.
Huh? What?
One important.
Are you asking because they switched from 20 to 16?
Yes. 20 to 14.
No. Somebody switches from 20 to 16, and therefore we lose share? No.
Thank you.
At least, not primarily.
All right. Next question will be coming from Goldman Sachs, Donald Lu.
Good afternoon, Chairman. I think it's getting quite complicated on this market share.
I know. It doesn't have to be so complicated.
I think-
I just look at the total. I just look at the hours.
Yes. I think maybe let's just look at the total TSMC's total foundry market share. I think last year we said you gained 4%. This year, based on your guidance.
Several points.
More than 4%.
Yeah.
How about next year?
Yeah. I think we'll at least stay there. After gaining almost 10 points, 4%, or not quite 10 points. Four points last year, several this year. That's more than halfway to a 10-point gain in two years. I think we'll stay there, at least. Just as you said, Donald, it doesn't have to be very complicated once you think about the basics. Yeah.
Okay. Next year-
Yeah. Mm-hmm.
TSMC will at least maintain market share-
in the whole foundry market.
Yeah. Next to you what?
Sorry. I'm just rephrasing your words. You've said next year, TSMC will at least maintain your total foundry market share next year.
Next year? Yeah.
Yeah.
We'll do everything we can to actually inch forward a little.
Okay, great. Yeah.
Remember now, actually, the looming battle is really not 16. In my mind, the 16 battle has already been fought. The looming battle with I don't know what the outcome of that battle will be yet. It's a 10. I'm not going to say anything about it today. Before I came here today, and after I leave here to go back to the office today, my thoughts are primarily on the 10, and not so much on the 16. The 16 is. Somebody said, Wellington, I think. The Battle of Waterloo was won on the Plains of Eton, and I have already graduated from Eton, and I can already foresee the results of Waterloo on the 16 there. The 10, I'm still in Eton. Okay?
maybe you can share with us, what are you thinking about 10? What's the key issues, maybe?
Maybe I didn't hear. Would you repeat?
My question is on 10 nanometer. You said you were thinking before and after the conference, maybe you can share with us like what you are thinking. What are the key points that we should be focusing on the 10 nanometer progress as well?
Donald's question is, what is in Chairman's mind regarding 10 nanometer?
What is in my mind regarding 10 nanometer?
Nanometer.
I'll shoot. Who knows what is in my mind?
I think right now, we reported our target for 10 nanometer, the development is in full speed in TSMC. At this time, our 10 nanometer is very competitive. As Chairman just mentioned, the particular intensity, that will be back on the historical trend even better. Most importantly is we have a few customers embrace the technology. Also the timing. Our leading customer even plan to collaborate with us to do their product tape out before we qualified the technology. Therefore, we see there is a strong market demand for the 10 at this point. We are very excited about this new technology definition. We're also very excited about our customer partnership of this 10 nanometer.
Yeah.
We are really focusing on the development. Yeah. Can I have another question here? My other question is more on the margin side. Maybe Lora can explain to us, like you commented that in Q2, there is inventory adjustment and also a
20 nanometer ramp-up cost. I mean, what's the total margin impact there?
It,
Second quarter gross margin.
I do want to go back a bit on that, I think someone from here asked a question about capital intensity.
Right
Where it go down. At that time, I said unhesitatingly that it will go down. I was just checking with Laura, because I remembered something, I think there is a chance that next year it actually may perk up a little bit, not significantly. This year, the general trend is very definitely downwards. We are already over the hill this year. This year, we'll be spending about less than $10 billion, or a little less. Last year we spent what?
9.7.
9.6. 9.6. Let's say that our revenue this year is 20-some% greater than last year, the same capital expenditure. Who asked the question, by the way?
Goldman.
Yeah. Right. Our capital expenditure this year is about the same as last year, our revenue this year is 20-some% higher than last year. I expect the total trend to be coming down. However, next year, we may see a little aberration. A little one. Since we're going to be taking you again in 2016 on the 16 nanometer, probably need to spend a little more capital.
Chairman, may I can make some comment on that?
Sure.
Chairman say that next year, maybe our capital intensity will be slightly higher than this year, but it's very safe to say it will be much lower than 2013.
Yeah. 2013.
Was 48%.
Yeah.
Lora, Donald still has a question for you on the margin of the second quarter. How much is the impact from inventory-
I'm sorry, did I interrupt something, Elizabeth? Yeah. Okay. Please. Yeah.
It was the inventory readjustment and the ramp-up of 20 nanometer impact to the second quarter margin.
We talk about the 20 nanometer impact first. In second quarter, we have very, very small shipment, but we have incurred some cost already. The impact to margin is about 1%, which point, okay, that's for 20 nanometer. For inventory adjustment, it's more than 1% but below 2%.
All right. Next question will be from UBS, William Dong.
Good afternoon, Mr. Chairman. We keep talking about technology. I guess the question I want to ask is that with all this rush to continue push down technology roadmap, to go down to 16, to 14, and to 10 nanometer, what are our thoughts about what's driving this demand? I mean, as we move toward, for example, internet of things, is there such a requirement to keep pushing on the technology front to actually have enough sufficient demand to keep driving it down?
Well, if the cost is low enough, cost is very much a part of the equation. If the cost is low enough, the demand will increase. We can see a lot of applications that are just waiting there. Of course, I'm talking about the mobile products, but I'm also talking about internet of things, wearables and so on. Internet of things. The applications are just waiting there for better, for faster speed and lower power, and higher-density IPs. Cost is definitely in the equation. Yeah, when you ask will the demand be there? If we can get the cost down to an acceptable level, demand will be there. Of course, that's how things like EUV come into the question. Nobody has asked about that yet. We actually were prepared to answer that. With the same answer that we gave you last time, by the way.
It's still a possibility to use EUV on one or two or just one layer in the 10 nanometer. Yeah, one layer. In 10 nanometer. Seven, I think, is of course, an even better candidate.
Thank you for that. It's good to hear that EUV will be potentially used.
EUV, we're working very hard with a lot of other people, ASML and other equipment vendors, to reduce the patterning cost of optical immersion lithography. We are also trying very hard to reduce the cost all around, not just the lithography, but all other costs.
Okay, thank you. I have one more question. As we look at the earnings continue to improve, and obviously expectation will continue to rise, what are your thoughts about potentially having more cash dividends as we move forward in time? Because right now, we've been staying about TWD 3 for the last few years, but as we generate more and more profit and cash flow improves, how likely is that to happen, for cash dividend to rise?
What?
As our earnings continue to improve, will we also increase our dividends?
Dividends? It will be seriously considered. An increase will be seriously considered.
Okay, thank you.
Yeah. We do have about TWD 7 billion worth of bonds that we have to repay, right? Certainly, an increase in dividends will be seriously considered.
Okay. I think we need to go back to the line for the next caller on the line. Operator, please proceed.
Thank you. Your next question comes from the line of Mehdi Hosseini from SIG. Please ask your question.
Yes, thanks for taking my question. Dr. Chang, going back to your commentary about 10 nanometer and how the tape outs is going to be available in the second half of 2015, that's when you're going to start ramping 16 nanometer. If I'm one of your major customers, why would I want to switch from 20 to 16, and why not just skip 16 and just go directly to 10? I have a follow-up.
All right, Mehdi, your question is with respect to what we said earlier about having 10 nanometer customer product tape outs available in 2015. Well, I think we said second half of 2015, not 2016.
Yes.
Your question is whether or not we see customers skipping 16 nanometer and going from 20 nanometer to 10 nanometer. Is that your question?
Yes, that's correct. Thank you.
No. Those customers are not skipping 16. I think there are segments of applications driven by the integration and possibly lower power also together to drive this technology. That's only a particular segment. That doesn't mean that driving a technology faster, the previous node will be short-lived. It's not, because the market application is spreading out. Some of the segments continue to drive the leading edge, but there are a lot of segment applications that stay on the older nodes. It's not a one move to the next node, yet the previous node will be idle. It's not the case. We see this migration fuel the semiconductor growth in those applications.
Thank you. This follow-up is a clarification. The capital intensity may increase next year, but does the dollar value of CapEx remain the same, or would it go up in 2015 compared to 2014?
Mehdi, your question is if our capital intensity ratio will go up a little bit in 2015, does that mean that our revenue will go up in 2015? Is that your question?
Well, more so, let me rephrase the question. During the last conference call, there was a comment that CapEx in 2015 would be similar to 2014. Is that assumption still staying, or should we assume that CapEx may actually go higher in 2015?
Okay. Will the comment of higher capital intensity in 2015 mean higher CapEx in 2015?
Yes, that's right.
Okay. Okay, now coming back to the floor, next question will be coming from Deutsche Bank's Michael Chou.
Chairman, one question for InFO. Do you expect any InFO revenue next year?
Oh, you mean the Eagle?
InFO.
InFO.
Oh, InFO. Okay.
Please do answer that question.
No, not yet. Thank you. The second question is regarding the 10 nanometer. Do you think the 10 nanometer total addressable market will be still bigger than 16 nanometer, 14 nanometer, or similar? Thank you.
Will 10 nanometer market be similar to 16 nanometer, 14 nanometer? I would be just guessing. I would just be speculating. If we talk about dollars, I think it'll be more. If we talk about wafers, I think it will be either similar or maybe a little less. Again, I'm speculating with you, all right? Because a lot depends on what cost we can come down to.
All right. Follow-up question from Dan Heyler, Bank of America Merrill Lynch.
Hopefully this question simplifies and doesn't complicate things. Just to make sure I understand this share loss thing. Basically, what you're saying is the share loss at 16 nanometer, these are customers that are choosing to skip 20 nanometer. Is that how I should think of this? Are any of these customers that are currently 20 nanometer that are going to 16 nanometer next year, or is this all people that are choosing to skip 20 nanometer?
first of all, I want to question the word-
Loss
share loss, okay? I don't consider it a share loss because just like 32, 28, we had real share in 32. We were very successful in 28. The two really belong to the same generation. 20 and 16 also belong to the same generation. Yeah. Share loss means that you start with something and then you lose it. It becomes less. This year, everybody has real share, okay? I'm just saying that we'll start on 16. We'll start with a lower share than we-
Contrarily
did with 20 or 28. We start with a lower share than we did with 20 or 28. We'll get back to a high share in 2016. That's arguing with him, but he did have a question all the same.
What is your question?
Well, the 20 nanometer. A simple question about the 16.
Oh, okay. Yeah. I remember. Dan's question is this lower share that we will have at 16 in 2015, is it due to a customer's
Just simply, are these customers moving to 16? Are these ones that have currently been on 20? Are these the guys that are skip-- Because the debate in the industry is should we go straight to 16 and skip 20? Are these customers that are basically have been at 28 and are skipping 20 and going straight to 14, your competitor?
Why do we start with a lower share? Well, mainly because Are you listening to me, Dan? I'm trying to answer your question.
Yes, Dan, you have to listen.
I'm listening.
Yeah. Mainly because our customers want it sooner. We got a little late, as I said.
Our customers want it sooner. That's why we're starting, and we'll catch up only a little later.
Very clear. Thank you.
Yeah. Mm-hmm.
I think we have a follow-up question on the call. Operator, could you please proceed to the next caller?
Thank you. Your next question comes from the line of Steven Pelayo from HSBC. Please ask your question.
Great. Thanks again for letting me come back with another question. I'm curious about customer concentration. In your annual filings, you had one customer, I think 22% or 23% of revenues. Seems like you're ramping, I think, another significant customer. I'm curious if you think about maybe just your top five customers in 2015, let's say next year, it seems to me that those five alone could maybe even approach as much as 45%, 50% of revenue. Do you think that's possible, and do you have any concerns about increasing customer concentration?
Steven, you are asking us about the customer concentration in 2015, and whether or not the top five customers that we will have in 2015 will account for about 45% or 50% of the total.
Just in the future. Doesn't necessarily have to be 2015, but even the second half of this year, it seems like you're going to be ramping up pretty significantly another customer, and so it would seem that your top four or five customers are going to be a very large amount.
What's this?
The customer concentration starting second half of this year, that our top customers will have a very large share of our revenue.
Is the customer concentration going to change? Is that the question?
It is.
It seems like the customer concentration is increasing going forward. Last year, you divulged in your annual filing one customer at 22% of revenues, I think. I think you have another customer that's ramping up very significantly now, and then maybe a couple other ones. That would suggest to me that maybe your top four or five customers are 40%-50% of revenue. Does that sound right? Is concentration increasing, and are you nervous at all of the risks of higher concentration, concentrated customer base?
Well, if you talk about our top five customers, I think that our top five customers have always had 40%-50% of the revenue.
Oh, okay.
Yeah. Now, something has changed in the last couple of years. To detect the change, you have to ask about the top three customers.
Okay.
Their concentration changed. Yeah. The top three customers currently has changed, but I don't see it changing very much from this point, which is the second half of 2014, in the next 18 months until the end of next year. I think the top three customer concentration is going to stay about the same.
Okay, thank you. One final question from me.
Yes.
I'm curious just about, you're pretty far, your view, from actual end demand, the data points on end demand in the first half of the year, maybe PCs were a little bit better than expected on end-of-life Windows XP. In general, some of the data that came out in TVs and smartphones wasn't that good, and there's a lot of expectations with this 4G ramp in the second half of the year. I'm curious, what signals do you have or confidence in that end demand will be robust enough to support so much, in my opinion, undifferentiated 4 to 6-inch blank slate smartphones out there?
Steven's comment was that it doesn't appear to him, I guess, that the smartphone demand in the second half will be that strong.
He does not-
Believe the smartphone demand-
Believe the smartphone demand-
in the second half will be that strong.
What does he mean?
I'm merely saying the data points we've seen thus far have been disappointing, so we need a pretty big ramp in the second half of the year.
I think we have got some numbers also. Anyway, he believes that smartphone demand will not be so strong, and therefore?
What drives the business growth in the second half?
What drives the business growth? Well, I guess we have the orders already, don't we? As far as third quarter is concerned, and the fourth quarter, we have some orders also. Well, I think maybe. Yeah, Mark?
I think Chairman did mention second half will go through some mild inventory correction. However, because of our share on 28 and 20 nanometer prompts us to have a stable trend.
Okay, fair enough. Thank you.
All right. Yeah. Coming back to the floor, there is follow-up questions from Barclays, Andrew Lu.
Morris Chang and Lora, I remember January conference, when the Q1 utilization low, we build the inventory for customer and, believing some of the inventory was being taken out. Second quarter seems doing very well, this new strategy. Assuming in Q1 next year, we have a larger decline because customer have adjustment inventory. Will we do the same thing in Q1 next year?
Yeah.
To smooth out.
I'm glad you praised it as a very good strategy. I thought so, too. I was the author of that strategy. All right. I thought so, too. It wasn't necessary for too long.
Yeah.
Very quickly, our inventory building program was overtaken by customer demand.
Will we do more this time compared to this year? Will be a kind of conservative.
Well, if your division is 100% just to respond to customer's demand, you have no capacity to build inventory. That's what happened when I said that program wasn't necessary for too long.
I see.
We didn't have the capacity anymore after, I guess, since March, maybe? Since March, yeah. That program was in place, inventory building program. Inventory building meant anticipating orders, not having orders and building inventory. That's what it means. That was the case, I guess, in the fourth quarter last year and in January and February. From March on, we began to be flooded by orders. There was no
No inventory.
Huh?
Show light.
No capacity to build the inventory with. All the capacity had to be devoted to building orders.
Yeah.
My second question, very quick. Earlier, Dr. Chang mentioned the ramp up on the 16 nanometer was starting from end of next year and change to second half next year. Can we just have a clear picture of what kind of revenue contribution will be by Q4 next year from 16 nanometer? Thank you.
What's the question?
How much will 16 nanometer account for our fourth quarter revenue in 2015?
16 accounts for?
Our revenue in 4Q next year.
How much? Say the year and the Yeah.
16 nanometer in fourth quarter 2015.
That's your question? Do you want to answer?
Yeah.
Okay, Lora.
We will start mass production of 16 nanometer from third quarter next year. It's bit too early to tell you the percentage of revenue contribution, but we believe it will be a single-digit range.
Okay. Follow-up question from Deutsche Bank's Michael Chou.
Chairman, regarding the 16/20 nanometer, could we say your total market share in 16 and the 20 nanometer will be similar to 28, 32 for the corresponding period? Can we say that?
Say that again.
Combined 20 nanometer and 16 nanometer market share, will that be similar to the combined 32/28 nanometer market share?
The combined 20 I just ran an analysis just a couple weeks ago, so I know exactly the answer to your question. The combined 20/16 market share in the first two years of its existence, which is this year and next year, well, I guess I have to add in 2016. The combined 20/16 share in 2014, 2015, and 2016 is going to be greater than our combined share of 32 and 28 in two thousand
2012. 2012.
Is it 2011 or 2012?
12. 12, 13, and 14.
Yeah. 2012, 2013, and 2014.
Thank you.
Great. All right. In the interest of time, I'm just going to allow the very last one, and that's Credit Suisse, Randy.
Thank you. This one might be more for Lora. Just wanted to ask on 20 nanometer, the forward look, how much impact on margin, how long it might take to reach corporate average, and if you could do an initial look on structural profitability and depreciation for next year.
Okay. I just talked about the second quarter, that 20 nanometer has about one percentage point impact to corporate gross margin. As we're ramping very fast in third and fourth quarter, impact will be larger. Based on my current number, it will be ranging from three to four percentage points in second half of this year. For next year, it will be 1% or 2%, roughly, okay? When to reach a corporate average? I think it usually takes seven to eight quarters, so it will be in 2016. I don't know which quarter yet. You asked about depreciation as well? We have not finalized our CapEx for 2015. I think the magnitude of year-over increase will be much, much smaller than the 33% this year. I don't have a firm number yet.
Okay. As you know that we are very confident in our profitability, we will end our conference call for this quarter right here. Before we conclude today's conference, please be advised that the replay of the conference will be accessible within three hours from now. Transcript will become available 24 hours from now. Both of which can be available through our website at www.tsmc.com. Thank you for joining us today. We hope you will join us again next quarter. Goodbye and have a good day.