Welcome to TSMC's second quarter 2012 earnings conference and conference call. This is Elizabeth Sun, TSMC's Director of Corporate Communications and your host for today. This is the first time that we're combining the quarterly earnings conference with the conference call. The event is webcast live via TSMC's website at www.tsmc.com. If you are joining us through the conference call, your dial-in line 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's Senior Vice President and CFO, Ms. Lora Ho, will summarize our operations in the second quarter and give you our guidance for the next quarter. Afterwards, TSMC's Chairman and CEO, Dr. Morris Chang, will provide his general remark on business outlook and state a couple of key messages.
We will open the floor to 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 www.tsmc.com. Please also download the summary slides in relation to today's earnings conference presentation. Before we begin, I would like to remind everybody that today's discussions may contain forward-looking statements that are subject to significant risk 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, good evening, and good morning to everyone. We had a very good second quarter. The financial results came in at the high end of each of the guidance. Revenue grew 21% Q-over-Q to set a record of TWD 128 billion. The strong demand for mobile computing devices and our leadership in 28 nanometer gave us a strong growth in the second quarter. On the margin side, second quarter growth margin was 48.6%, or 0.9 percentage point higher than that in the first quarter. The increase of growth margin mainly came from the higher capacity utilization across all technologies. Although 28 nanometer growth margin is currently below corporate average, we expect it will reach to corporate average by first quarter 2013. Operating margin was 36.5% in second quarter, up 2.9 percentage points.
Both R&D and SG&A expense as percentage of revenue decreased by about one percentage point each. You may notice that we had a loss of TWD 0.8 billion in the non-operating items. This mainly due to a one-time impairment charge of TWD 2.68 billion on our 5.6% holding in SMIC shares. This contributed to a TWD 0.09 drop of our second quarter EPS. Overall, our second quarter EPS was TWD 1.61. ROE was 26%. Let's move on to revenue by product segment. We have seen revenue from all applications increase sequentially. Among the four major product segments, communication increased by 27%, computer increased by 20%, consumer increased by 9%, and industrial-related revenue increased by 39% in the second quarter. The high growth of industrial and standard application is mainly due to strong demand for ICs used in mobile computing devices, such as power management IC and a touch controller.
In terms of technology, revenue from 28 grew nearly 90% in the second quarter. We expect shipment of 28 to double in the third quarter. Revenue from advanced technologies, that is 65 nanometer and beyond technology, accounts for 61% of our second quarter revenue. Take a look at the balance sheet. The cash and marketable securities ended the second quarter at TWD 188 billion. Accounts receivable and inventory amount went up as a result of business growth. Current liability increased by TWD 86 billion, mainly due to the accrual of dividend payable of TWD 78 billion. On the cash flow side, we generated TWD 70 billion from operations. Invested TWD 59 billion in capital expenditure, repaid TWD 3.9 billion in short-term loans. As a result, our cash balance increased TWD 7.6 billion to TWD 178 billion at the end of the second quarter.
Free cash flow generated in the second quarter was TWD 10.5 billion. Let's look at the capacity. Our Fab 15 began volume production of 28 nanometer in the second quarter, and we expect to ramp at a faster pace in the second half of this year. Given our CapEx, we expect our total capacity to increase by about 14% year-over-year, and 12-inch capacity will increase about 21%. Let me provide you our guidance for the third quarter. Based on our current business expectation and a forecast exchange rate of 29.76, we expect our revenue to be between TWD 136 billion and TWD 138 billion, which is a sequential growth of 6%-8%. We expect the third quarter gross margin to be between 46%-48%, and operating margin between 34%-36%. This concludes my remarks. Let me turn the podium to Chairman.
Thank you, Lora, good afternoon, ladies and gentlemen. Today, I will make a few comments on second quarter and third quarter, I'll also give you some color on world economy, supply chain inventory, and our fourth quarter outlook. I will talk about a few major technologies of ours. Lastly, I will talk about CapEx, capital intensity, and growth. Second quarter was a good one. We were actually quite pleased with it. This year, every quarter, the major effort has been to ramp up 28 nanometers. In doing so, of course, we did incur a lot of costs. Also, as a result, the gross margin of 28 nanometer all year this year will not be up to the corporate average standard.
In spite of all that, also in spite of an extraordinary item in the second quarter, which Lora mentioned, the impairment charge of SMIC shares, which actually accounted for TWD 0.09 earnings per share. In spite of the unusual costs in the 28-nanometer ramp-up and the unusual item relating to the impairment charge of the SMIC shares, second quarter was very good, we expect a good third quarter. We'll see a growth, as Lora has already guided. At the midpoint of our guidance, we'll see a growth of about 7% in revenue in the third quarter, sequential quarter-to-quarter revenue growth of about 7%. Since we will not have the impairment charge in the third quarter, our EPS growth between second and third will actually be stronger than 7%. It looks okay.
As we look further into the future, fourth quarter and the first quarter of next year, we do have some worrisome signs. World economy, as you know, and I will not draw on it, the outlook. I'm actually talking about the outlook, the future outlook of world economy, and I'm comparing it now with the outlook as most people saw it four months ago, three months ago, or six months ago. The outlook now certainly has deteriorated in the last three to six months. The U.S., which matters the most to us because our market is still very much, is majorly in the U.S. The U.S. economy matters to us the most. It has gone into a less optimistic situation than we saw it even three months ago. The very good job creations record early in the year has now disappeared.
There does not seem to be any political solution in sight for the forthcoming financial cliff at the end of the year. The retail sales, recent data are not good. The outlook for the U.S. economy has deteriorated in the last few months. Next to importance to us of the U.S. economy, you have Europe, Japan, Mainland China, and Taiwan. I would say that the outlook for any of those economies has not improved in the last three months. We, of course, do pretty thorough, pretty extensive market research on our business. One key factor, besides the world economy, one key factor, of course, is the supply chain inventory of our products. That is not good. I'll give you some numbers.
At the end of the first quarter, the overall supply chain inventory in days of inventory at the end of first quarter was six days below seasonal. At the end of second quarter, it was three days above seasonal. We are forecasting that at the end of third quarter, it will be 12 days above seasonal. This, of course, indicates that there will be a correction in the fourth quarter. We indeed are forecasting that there will be an inventory correction in the fourth quarter to about at the end of fourth quarter, we forecast that the inventory will be only eight days above seasonal. 12 days at the end of third quarter, 12 days above at the end of third quarter, and eight days above at the end of fourth quarter. We also look at our customers' days of inventory.
The fabless customers at the end of first quarter was two days below, and at the end of second quarter was four days above. We are forecasting that at the end of third quarter, customers' DOI will be 10 days above, 10 days above. There will be a correction, but at the end of the fourth quarter, it will still be six days above. Those numbers are for the fabless customers. For IDM customers, the pattern is similar, but numbers are different. The pattern is similar, so I'm not going to talk about them. Our fabless customers accounted for a very large, the dominant majority of our sales anyway. We are now seeing a dip in our revenue in the fourth quarter, a dip from the third-quarter level. I think it will be a dip in our forecast, not just I think.
Our forecast is that the dip will not be nearly as serious as the dip we experienced in the fourth quarter of 2008. That I wouldn't even call a dip. I would call that a plunge. This one, I think, is a dip of certainly far more modest magnitude than that one. As to exactly how much it will be, I think it's too early to say. We can see that there will be a dip. We have, of course, accordingly made preparations. We have seen it coming for, oh, at least a month, maybe two months now. We saw early signs of it actually three months ago, but those were very early signs. Two months ago, we became surer, and one month ago, we became pretty sure that there will be a dip in the fourth quarter.
Further, we also think that the dip will continue into the first quarter. Still further, we see a pretty healthy recovery in the second quarter. In summary, I'm saying that we will have a dip that will last two quarters, fourth quarter and first quarter next year. By the second quarter, it will have rebounded pretty strongly. Those are the indications that we have now. I'd like to say a few words about our technology progress. 28 nanometer. It's progressing very well. Our output and our yields are both above the plans that we set for ourselves and the plans that we communicated to our customers early in the year. Early in the year means January, February of the year, we set our plans in output and in yields.
We, of course, ever since then, we tried to exceed the plan, and we had also communicated the plan to our customers at that time. We have indeed exceeded the plan in both output and yields. We expect to ramp up to about 68,000 wafers per month by the end of the year, 28 nanometer, 68,000 12-inch wafers per month by the end of the year. By fourth quarter, we will be nearly caught up with the demand now, and we expect to fully meet the demand from the first quarter on. First quarter of 2013 on, we'll fully meet the 28-nanometer demand. It is also then that we expect that the 28-nanometer gross margin will catch up with the corporate average. As I said today, both the defect density and yields are better than 40 nanometer at the same stage of the volume ramp.
They are also better than what we planned early in the year and what we communicated to our customers at that time. Next, a few words on 20 nanometer, 20 SoC. We have made very good progress on the 112 megabit SRAM unit. There are still challenges to overcome in meeting our yield plan of the entire chip. We have made very good progress on 112 megabit SRAM. There are still challenges to overcome in meeting our yield plan of the entire chip, which has both the logic and the SRAM on it, of course. Our 20 nanometer SoC, we believe is fully competitive with industry leaders, other companies, 22 nanometer for the served available markets that we serve. For our markets, we believe our 20 SoC is fully competitive with anyone's 20 or 22 nanometer offering.
One important point to make is that our 20 nanometer has the industry's leading metal pitch of 64 nanometers. Our leading competitors have 80 nanometer metal pitch. That allows an advantage in the device's density and die size. Now, as for the timing, we expect our 20 nanometer technology to be qualified by the end of this year, and we'll be ready to support customers' tape outs in Q1 of 2013. Now today, last time, I mentioned that we will have a FinFET product after 20SoC. Today, I'm glad to say that we have been planning the 16 nanometer FinFET. Right after our 20 nanometer planar, which is the 20SoC, we will offer FinFET at 16 nanometer for significant active power reduction. We expect to achieve speed and logic density levels comparable to industry's leading players' 14 nanometer FinFET.
We expect our 20SoC to be competitive with competitors' 22 or 20 products. We expect our 16 FinFET to be competitive with our competitors' 14 nanometer FinFET products. You might ask why are we calling it 16? The only reason, in fact, until 2 days ago we were undecided on whether to call it 14 or 16 FinFET. Now, the only reason we decided to call it 16 FinFET is first, we want to be somewhat modest. Second, we had told quite a few major customers of ours, the 16 FinFET, that designation, and we didn't want to confuse our customers by now switching to 14. We expect it to be competitive with other people's 14 offerings. Now, the 16 nanometer FinFET, our 16 nanometer FinFET, is expected to deliver about 25% speed gain given the same standby power over the 20 nanometer SoC.
It's expected to give 25%-30% power reduction at the same speed and the same standby power. For mobile products, it's expected to give 15%-20% speed gain at the same total power. As for timing, we expect it to be about 1 year after 20SoC, namely, it should be ready for risk production at the end of 2013 or early 2014, about 1 year later than the 20SoC. Now, I want to make some comments about CapEx, capital intensity, and growth. I know that several analysts have written about foundry industry's capital intensity and our TSMC's capital intensity, and so on. I'm addressing this subject today because I have seen all these reports. I've seen them without agreeing with them. Without agreeing with some of them, anyway. There are some that I do agree with.
Now, why are we having such high capital intensity now? I think this is actually a focus point of our internal discussion among our top-level managers for the last 2 years now. Basically, we invest in capacity to get future growth. You look back at history. If you look at our TSMC history, during 1997 and 2002, between 1997 and 2002, during that 6-year period, TSMC's capital intensity ratio stayed mostly above 60%. 60% during that 6-year period. As you recall, there was a high-tech bubble bursting in late 2000 and early 2001. In spite of that, our revenue CAGR between 1997 and 2007, here, after having spent a lot of capital, having sustained high capital intensity for 6 years, 1997 to 2002.
Our revenue CAGR between 1997 and 2007, that's a 10-year period, was 20%, compounded annual growth rate of 20% in revenue during the 10-year period. The first six of which was marked by high capital intensity. During that 1997 to 2007 period, foundry industry growth was 16% in the same period. Ours was 20%. As a result, our foundry market share rose from 31% in 1997 to 43% in 2007. 31% in 1997 to 43%, 2007. When we had those internal discussions about capital, and that's really a major focus of our internal discussions, top-level managers, I mean. We look at four things. First, are we going to be the technology leader? In the capacities that we are investing in. The first question is, are we going to be the technology leader?
The second question we ask ourselves is, are we going to be able to retain our leadership in flexible and responsive manufacturing? We ask our question, technology, are we going to be the leader? Manufacturing, are we going to be leader? Third question that we ask ourselves, are we going to retain the customers' trust, major customers' trust, or perhaps are we going to even add customers? The fourth question we ask ourselves is, at the price and cost we expect on the new technologies, the capacities of which we are investing in, at the expected cost and price, are we going to be profitable? Are we going to be able to make the same kind of money that we have gotten accustomed to?
Only if the answers to all four questions is yes, only if we are confident in those four issues, four points, do we start to spend the capital money. I actually have not made a secret of that. In 2010, I believe late 2010, I told you that 2010 was the first year we started to spend a lot of capital. In late 2010, I told you that our goal was, in the following five years, achieve a pretax profit growth of 10% CAGR in the 2010 to 2015 period. At that time, which was late 2010, I said that our financial goal was to achieve a 10% pretax income growth CAGR and retain or exceed an ROE of 20%. 10% pretax income growth, 20% ROE. That was two years ago, 2010, we have now raised our goals.
It is not 10% anymore, although the ROE, we still keep it at equal or greater than 20% ROE. The pretax income growth goal is more than 10% now. I'm not prepared to answer you yet what it is. Let me assure you that when I say it's more than 10%, I don't mean that it's 10.1%, okay? It's significantly above 10%. We believe that this is not only the right strategy, it is the only strategy. If you want to do well by your shareholders, we believe it's the only strategy. As far as this year's CapEx is concerned, at this point, we are still following the guidance that we gave you last time, I believe, $8 billion-$8.5 billion. At this point, we are still following that. Next year, we are not going to forecast until early next year.
I think I have already given you a view of our reasoning and our strategy and our objectives. As to the exact number, I will not give you until early next year. All right, I believe I have finished my prepared comments. I believe we are open for questions, are we not?
Yes. This concludes our prepared statements. Before we begin the Q&A session, I want to remind everybody to please limit your questions to two, no more than two at a time, to allow all participants an opportunity to ask questions to the management. Questions will be taken from the floor as well as from the call. Should you wish to raise your question in Chinese, I will translate it to English before our CEO or CFO answer 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 which they were received. If at any time you would like to remove yourself from the questioning queue, please press the pound or the hash key. Let's begin the Q&A session.
Our first question comes from the floor, and that goes to Bank of America Merrill Lynch, Dan Heyler. Use the mic.
Thank you very much, Elizabeth. Thanks for the new format. Hopefully, we'll all get a little more sleep. I hope you're feeling better. It sounds like you have the same cold that I have, Dr. Chang.
Yeah.
Quick question. On the IDM models here, we've seen the IDM model work for high volume businesses such as the CPU business and the memory business. Given the huge amount of demand and growth in the application processor market that we're seeing proliferate in the mobile area, and with competitors scaling up their manufacturing facilities, I'm wondering if it would help TSMC's efficiencies to start to dedicate some lines or specific fabs to be more product-focused as you go forward in these very high volume businesses, or will you keep your very large broad-based fabs?
The question is, are we going to dedicate more lines to specific products? Is that right?
Will your manufacturing strategy change? Yes.
No. You first started to talk about IDM and were you asking me about the future of the foundry fabless model? You wrote about that also.
Do you think you need to dedicate some fabs to product-specific areas that are very high volume?
Actually, yes. I think that's almost a natural outcome the way the market is trending. I think that there are going to be larger customers. It makes complete sense to dedicate a whole fab to just one customer. A whole fab or two whole fabs, in fact, to just one customer. Now, remember, we made our mark in serving many customers. In fact, that's really part of our secret sauce of success. The ability to serve many customers to their satisfaction. We'll retain that capability. There are customers that are getting bigger and bigger, so it makes sense that we dedicate a whole fab or even more than a whole fab to just one customer. As far as specific products are concerned, well, right now we are already concentrating.
For instance, Taichung will have the vast majority of 28 nanometers, whereas Tainan will have the vast majority of 20SoC and 16 FinFET. Both of those manufacturing centers are under one manager. Taichung is under one manager, Tainan is under one manager. Yeah. I don't know whether I answered your question.
Yeah, you did. Thank you very much. Then I'll get back in the queue. We heard you today, as well as ASML, talk about 20 nanometer ramp towards- I think tape outs in next year, in 2013. We're seeing obviously some significant challenges currently, as you highlighted in 20, 28 nanometer with High-k metal gate. Given that you've got High-k metal gate challenges, double patterning on 28, two big changes, what's the visibility in your sense in really being able to execute 20 nanometer in the second half of next year? Would we be able to see volume there, and what gives you the level of confidence?
I think that we'll start some production of 20 nanometer next year. Small scale. Very low. What we would call risk type of production next year. 2014 will be a ramp year for 20 SoC.
Okay.
We are pretty sure of that.
Just a quick clarification. Earlier comment, you had talked about FinFET, that you would be competitive at 20 nanometer-
Well-
with your-
Yeah. In answering you, I haven't.
Oh, go ahead.
included it. 20SoC, which is planar, will ramp in 2014. We believe that the 16 FinFET will ramp in perhaps the second half of 2015.
Okay, great. Just one clarification on something you said, if I may. You talked about TSMC being competitive at 20 nanometer relative to the industry leader, who's at 22.
Competitive.
Competitive, right. That competitor, I believe, is doing FinFET at 22. Are you including that in your statement? Okay. Thank you.
I also said in our served markets. Yeah.
True.
Yeah. Which would not include high performance CPUs.
Sure. Mobile. Thank you.
Okay. All right. Since we have people on the conference call, I'm just going to open the line to the call first. We'll take our next question from the call. Operator, please proceed with the first caller.
The first question today comes from the line of Mehdi Hosseini from Susquehanna. Mehdi, your line is now open.
Yes. Thanks for taking my question, thanks for the new format. Two question. Dr. Chang, you talked about the Q4, Q1 trend. At the same time, the 28 nanometer gross margin should reach the corporate average. How should we reconcile lower shipment as a result of customers reducing inventory with a better margin profile for 28 nanometer? I have a follow-up.
The question is why couldn't we delay shipments until the margin becomes better? No?
I think.
No.
Because.
As you mentioned, a dip in shipment in Q4, Q1 time frame, and that obviously will have an impact on utilization rate and margin. How should I reconcile that with 28 nanometer margin profile that is actually improving and reaching the corporate average?
Mehdi, your question is Q4 and Q1 will appear to be a down quarter, where we will have some margin pressure. At the same time, our 28 nanometer margin will go to the corporate level by first quarter. How do we reconcile these two?
Well-
Yes
The way to reconcile those two is that 28 nanometer will only account for about 20% of our revenue in the fourth quarter this year, and it will account for a little more than 20% of our revenue in the first quarter of next year. While the 28 nanometer gross margin will be climbing, is climbing, and will be climbing, the rest of the products margin will drop because of lower utilization.
Got it. My follow-up has to do with the 28 nanometer and 20 nanometer capacity for next year. How should we think about the additional 28 nanometer capacity compared to the 20 nanometer pilot line? Do you have any thoughts on how aggressive you want to be? At the same time, you want your customers to try out the 20 nanometer. I'm just kind of confused how the product portfolio for those two nodes are going to converge.
You are asking us about the capacity plan next year for 28 nanometer as well as for 20 nanometer. That's right?
Yes.
I would just-
Especially as some of the 20 nanometer may move to 20 nanometer.
Let me just describe our capacity plan in the following way. This year, we will be spending between $8 billion-$8.5 billion in capital CapEx. About $1 billion-$1.5 billion will be spent on 20SoC, 20 nanometer. I think around $6 billion will be spent on 28 nanometers, and the rest, just odds and ends, including R&D. This year, the vast majority of the capital spending is still on 28, but 20 nanometer has already made a significant appearance in CapEx. Next year, there will still be some capital spending on 28 nanometer, but relatively small, and the vast majority will be on 20 nanometer. That kind of spending, that pace of spending on 20 nanometer will continue into 2015. Of course, in 2015, the 16 nanometer FinFET will also be making an appearance.
Fortunately, the conversion from 20SoC to 16 FinFET is quite good. In other words, we don't expect any significant loss in the conversion from 20 nanometer capacity to 16 nanometer capacity.
Got it. Thank you.
All right. Now we will switch back to the floor. The first question goes to Deutsche Bank's Michael Chou.
Hi, Chairman. One question is, would you invest in ASML EUV, given that the industry leader is going to invest as a EUV? What's your view.
Which industry leader?
As you mentioned before. Yeah. What's your view for the competition between ASML EUV and Nikon's multi E-beam? Which one will become-
I'm sorry, I didn't get that.
Okay.
You were asking about ASML?
ASML.
What you said competition or-
The competition between ASML's EUV and Nikon's multi E-beam methodology. Which one would become the industry standard going forward?
Well, which one will become. Let me put it another way. It appears that the multi E-beam is behind EUV. EUV progress has not been very good either. We're still going to need the EUV even though the progress to date has not been very satisfying. If you compare it with the E-beam, I would say that E-beam is certainly behind EUV.
Does that mean, are you going to invest in ASML's EUV going forward given that your competitors move?
We are actively negotiating with ASML. Actually, ASML brought up this investment R&D deal to three companies together, the three industry leaders together. Now, of course, one of them decided to do it first. That's okay. We have been for more than half a year now, and recently, of course, since one of our colleagues has already signed. Of course, that kind of got our attention again. Our discussions with ASML have become even more active recently. Yeah. We are still in active negotiations with ASML.
Thank you.
Our next question goes to Morgan Stanley's Bill Lu.
Hi, Dr. Chiang. You just raised your pre-tax income CAGR from 10% to something more than 10%. Can you talk about what is behind that? Is it a higher expectations now for market share? Is it the whole industry that you think is going to grow faster? Is it profitability? What exactly is behind that more bullish outlook? Thank you.
There are two main things behind that. One is that our lead in both technology and manufacturing, I believe is strengthening. Has strengthened. Remember, starting in 2010, we didn't just increase CapEx, we also increased R&D. Our R&D right now is double what it was in 2009. It was a two-pronged thrust back in 2009 and 2010. The two-pronged thrust was to increase both R&D and capital significantly. R&D is now double what it was in 2009. At any rate, I believe that the reason, two things behind why we raised our pre-tax income growth goal. One is that we believe that our technology lead has strengthened. We have maintained our manufacturing lead, which we have had all along. Our customer trust lead, which we have had all along.
The other reason, of course, is that the handheld products, the mobile products, the smartphone, and the tablets. That was something that we did not completely foresee in 2010. In 2010, we did not foresee this mobile products market, not as clearly as we do now anyway. Those are the two reasons why we raised our goal.
Great. Thank you. My second question is more short term. You talked about this dip or inventory correction in 4Q and 1Q of next year because of the macro factors and such. That to me feels like last year when the macro got a little bit worse. You saw a little bit of maybe a two-quarter period where you were growing below seasonal patterns. If you look at this year versus last year, are they similar? Worse or better?
You are right. It's very similar. I think it's a very similar situation. With some difference. I think that the European situation was certainly, I think this year we are worse than last year. Mainland China, I think, last year, we're talking about a slowdown from 10%-9%. Now this year, we're talking about a slowdown from 9%-7.5%. Basically, it fuels the high hopes early in the year, and mainly those high hopes are based on general economic progress. Early last year, there was high hopes about world economy also. Then the hope was dashed later on in the year. Now early this year, there was high hope again. It's being bashed now. Yeah. The inventories, I think were based on that also. The high hopes gave rise to the high inventories in the supply chain. Everybody hoped.
Everybody had high hopes. Yeah.
Okay, next question is Barclays, Andrew Lu.
Hello
You're okay, Andrew. No. Can't you make his microphone-
Hello?
Yeah, it's working.
It's working. Okay.
Yeah. Okay. Yeah.
Dr. Chang and Lora, two questions.
First one is, are we going to see a double-digit decline in any of each Q4 or Q1?
Double-digit decline in what?
Revenue.
You mean, what time period are you talking about?
Q4 or either Q1.
A sequential double-digit-
Yes
decline? As I said earlier, I don't really want to predict so early. Right now, we are looking at something that's in the gray zone between single digit and double digit. Okay?
Thank you very much. Thank you.
My goodness, you forced me to give you an answer. All right. Don't blame me if it turns out to be much better than that.
Yes. Always possible. Second question I have, I actually did some calculation. Assuming due to all these 20A, 20, 16nm FinFET investment continue, we might remain to see the capital sales ratio remain at 50%. Plus, I actually calculate the cash gifting, if we stick at the TWD 3 per share, that will take out about 15%-16% revenue as well. Total combined is about 65% revenue as a regular basis cash outflow. Our EBITDA margin, EBITDA divided by revenue is about 60%. Each year, we are going to have a 5% short on revenue as cash. This doesn't include the potential we might invest ASML.
I see.
What's our financing plan through the equity and debt for the next 5 years, based on these changes? Thank you.
Lora, will you relieve his concern?
Andrew, there will be few years that our free cash flow may not be good or may not be sufficient to pay the TWD 3 dividend. Since we have
Yeah. We're going to keep the TWD 3 dividend.
Okay. Since we have quite strong balance sheet, we have started to borrow by issuing some corporate bond, we're starting from last year. This time frame, interest rate is very low, so we were able to get a 1.3%, 1.4% type of interest rate for 5 year or 7 year. We will leverage that for the period that we will have a very high capital intensity. We believe when the revenue catch up and profitability, cash flow catch up later, and we will be in good shape. We're not too worried about that.
Working mainly on the debt financing?
Yes.
No equity raising?
No. No plan for equity.
Thank you.
Debt financing actually the latest ASML deal sort of Made me think also. There are other novel ways, innovative ways, and so on. I'm not saying that we'll do it. The answer to your question is, debt financing, no equity, maintenance of at least a TWD 3 cash dividend. Now, in addition to those definite answers, maybe there are innovations which we haven't decided on yet.
Next question goes to Citigroup's Roland Shu.
Good afternoon, Dr. Chang. Two questions for me. First, since now we are talking about the 16-nanometer FinFET. My question is, what next to 16? The 14 nanometer still on your technology roadmap? Or after 16, maybe we will move to maybe 11 or 10?
10, maybe. No.
Is 14 nanometer still in your roadmap?
Pardon me?
Is 14 nanometer still on your roadmap?
14?
Yeah.
I don't think so.
Okay. That means for TSMC
16. Our 16, we believe, will be competitive with other people's 14. Mm-hmm.
Yeah. Understood. Have the 14, and so how is your EUV? When is your EUV released?
What is our EUV?
EUV, yeah. I think that will be introduced as what kind of the technology? No?
I think it will be coming at 10, as our 10.
My second question actually is similar to Andrew's question. I think that since given TSMC now, we have a heavy invest on the 20 nanometer, next year 20 and going forward 16, and also we have EUV on 10. My question is, are TSMC considering to invite your key customers to invest in TSMC, like what ASML is doing now, invite a technology leader to invest ASML.
Oh, I must clarify the careless comment I made earlier when I talked about innovative things such as we are not considering. Actually, we've made a definite answer to Andrew's question. We are not considering any equity offering at all.
Okay.
We're not considering equity offering, not to a customer, not to investors, no.
Thank you.
Yeah.
We will now take our next question from the call. Operator, please proceed with the next caller on the line.
The next question on the line today comes from Brett Simpson from Arete Research. Brett, please go ahead.
Yeah. Thanks very much. I have a question for Dr. Chang around 20 nanometer. We've seen Intel recently and several of your customers talk about concerns over the transistor cost at 20 nanometer. It's not falling like it has in previous nodes, at least that's their perspective. Because of the number of process steps that are increasing at 20 nanometer. I wanted to get your perspective on this. What do you think this means for the economics of the established business model? To what extent, if the costs are going to be rising at 20 nanometer, can these costs be passed on up the supply chain? Thank you.
Yeah. Can you hear?
20 nanometer transistor cost.
20 nanometer transistor cost. 20 nanometer transistor cost. The capital intensity has also introduced a pretty high component of depreciation cost. It has raised that component, depreciation cost, in the advanced technologies. Now, here, of course, the FinFET does have an advantage, at least ultimately. We are going to the FinFET in 2016. For 20SoC, we do have the advantage of our denser metal pitch that I talked about earlier. The denser metal pitch resulted in smaller die. Even though the transistor cost might be higher, with a smaller die, smaller chip, the economics works out very competitively. It's actually a pretty involved technical calculation. That's the conclusion. The higher transistor cost, which is mainly because of the higher capital intensity, is compensated by the higher density. That's basically the answer.
Got it. Thanks. Thanks very much. Just a follow-up for Lora. Lora, can you perhaps talk about the relationship between depreciation and CapEx as we sort of go through this higher level of capital intensity? Today there seems to be a big gap between depreciation and CapEx versus history. How does this really trend over the next couple of years? If you can maybe just give us some help there, that'd be great.
Brett, your question is, how would the trend be given the high capital cost? What's going to be the ratio between depreciation and CapEx for the next few years, right?
Yeah.
The ratio.
Ratio between what?
Depreciation and CapEx.
Ratio between depreciation and CapEx. Well, actually, every dollar of CapEx equipment, every dollar spent on equipment is depreciated over five years. Every dollar spent on facilities is depreciated either over 10 years.
10 years.
Yeah. There are some that because
Yeah, for the building, it's depreciated over 20 years. They have 20 year for the building, 10 year for facility, and five year for equipment.
Well, of course, about 80% of our CapEx is on equipment, I think.
Exactly.
You asked about the ratio between CapEx and depreciation. Take 80% of the CapEx, depending on what point in time in the year it's spent, then you spread it over five years. That's 80% of the CapEx. I think roughly 20 year depreciation stuff.
Building.
Yeah, I know. It varies from year to year. This year, next year, we're actually building quite a few buildings. There is a greater component in our CapEx that's going to be depreciated over 20 years. Mm-hmm.
I think your question is more to that. I can give you one example for this year. You're asking the relationship between depreciation CapEx. For example, this year we were planning to spend $8 billion-$8.5 billion, with the depreciation about a little bit more than $4 billion. There's some relationship between the two. Going forward, actually it will depends on when you spend the depreciation, when you spend the money, which quarter, on what technology. We also have some 12 inch coming down from depreciation. Every year the number is different. That's very difficult to answer, to give a very simple answer for your question.
Okay, thanks very much.
Okay. We will continue to take our next question from the call. Operator, please proceed with the next caller.
The next question comes from the line of Mahesh Thangamany from RBC Capital Markets. Mahesh, please go ahead.
Yes. Thank you very much. Dr. Chang, I have a question on the 16 nanometer. What is your confidence that you can accomplish that without the EUV? Also related to that, if ASML supplies a stable machine, how long will it take for you to put it in production? I'm pretty sure you have 20-year problems to solve, like with the reticle, and the resist, and the line. How long will it take to solve those problems to put it in production?
Could you repeat the question?
Mahesh? Your question is, if we start with 16 nanometer, I suppose you are referring to 10 nm because that's when we will start using the EUV.
I think his question was, how confident are we that we can accomplish the 16 nanometer without EUV.
Oh, okay.
Isn't that right?
Yes.
Without.
Yes.
Yeah. Mm-hmm. Well, the answer is yes. We are quite confident. We are very confident we can accomplish the 16 FinFET without EUV. Mm-hmm. Now, it's the second question that I didn't completely get. He talked about radicals and all that.
you say, if we get a machine from ASML on the EUV, how soon will we begin the production? How soon will we be able to put the machine into production? Is that your question?
That is correct, because I assume that there are multiple problems to solve once the machine is there also in the fab. how many years it will take to put it in production?
after we get a machine, a EUV machine, how long will it take to get it to production? Is that the question?
Yes.
Well, we have had one machine for a year already. I can't tell when we will be using it in production yet. I think that actually, the fact of the matter is that by the time we use an EUV machine, which I think is in our 10 nm generations, I think by that time, this machine that we have had for a year will be obsolete.
Yeah.
We'll be getting new machines. As to how long it will take to get a new machine, you can tell that. I will tell you about the immersion machines. We've been using immersion machines for years. Sometimes it doesn't take very long. Sometimes it takes several months or half a year.
Okay. That's very helpful. Just one more follow-up on, can you give us your estimate of where your competitors are on 28 nanometer production?
There's a lot of rumors about, but I do not believe most of those rumors. I really haven't seen anything real yet. Well, I've seen very little real yet.
Okay. Thank you very much.
All right. Now we will switch back to the floor. The next question go to HSBC, Steven Pelayo.
Thank you. Very impressive performance on your industrial business, up nearly 40% quarter-on-quarter, 22% of revenues. That's now bigger than your computing segment. That growth rate's actually bigger than your 40 nm and below. I want to understand the outlook for this business. Is this sustainable? Is this a step function higher, and now sustainable? What are you thinking now for that industrial growth?
A pretty large part of it goes into smartphones and tablets. We believe in the growth of those. Lora, you mentioned already, you told everybody touch control and all that stuff.
Those MCU, data converter, flash controller, touch controller, those type of things.
Voltage. Yeah. Power. Yeah.
Power ICs.
Yeah.
Power.
I guess the question is, these are new businesses for you, there should be a level of sustainability to them, or is this still a cyclical thing and they're going to fall off as well?
I thought I was answering that question. I mean, mobile products are sustainable, aren't they? Yeah.
Just maybe one more follow-up on the competitive landscape.
Is it cyclical? I think everything is cyclical.
Sure.
Except maybe bread and rice.
If I could just follow up on the competitive landscape. We heard just this morning from Qualcomm looking to qualify more competitors. You talked about how you've raised your guidance for pre-tax profit. You were surprised at the strength of smartphones and tablets. Those customers are also looking for alternatives as well. The last question was really just specific to 28 nanometer, but I want to ask in general, the competitive landscape, do you feel that it is becoming more intense and now you need to be much more aggressive in pushing 20 and 16 nanometer? Or do you think the competitive landscape is more of the same?
Competitive landscape has changed because competitors have changed. Three years ago, two years ago. Anyway, when you talk about competitive landscape, you and I will both think of UMC, et cetera, with, at that time, GlobalFoundries emerging. Now you ask me what the competitive landscape is. Competitive landscape is Intel, Samsung, GlobalFoundries, UMC. Now two or three years ago, GlobalFoundries and the UMC were almost the two only ones. Now they are the two less important ones. All right, you asked me what the competitive landscape is. I really think that you know the answer. I actually read the same thing that you do. Maybe you read even more than I do. I mean, last thing I was reading was. I haven't finished reading yet. It was Intel's call. The transcript. I read all those damn transcripts. I'm sorry. Yeah.
Yeah. I think it's a very competitive environment. Very competitive. Our competitors have changed, and they are even more powerful and more intimidating than our own competitors. Not that we are intimidated. Okay.
All right, for the interest of time, I'm just going to allow one last questions from the floor, and that's Credit Suisse, Randy Abrams.
Okay, thank you. In the prepared remarks, you mentioned you're taking preparations for the dip. Could you talk about some of those preparations, if it's any change in spending or plans?
Preparations for the debt?
For the dip. The fourth quarter dip. You said you're taking some preparations.
Preparations. We have already had two rounds of cost reduction in the last two months. The first round, I initiated and I made it responsibility of every fab manager and every functional manager of the company. That, I guess, was about 30 or 40 of them, and each of them had a unit as for cost reduction. The results were compiled, and it was a pretty significant round. Just two weeks ago, I called for another round of cost reduction. Clearly, our objective is to keep the gross margin and operating profit margin up as high as possible. Preparations meant rounds of cost reduction, cost and expense reduction, COGS and operating expenses reduction. Yeah.
The follow-up question on the gross margin guidance, you guided for a small decline in the third quarter on rising sales. If you could maybe talk about the factors in the margin decline and maybe from these cost reductions, how we should think about OpEx growth the next few quarters.
Lora or Sun, yeah.
Randy, if you look at our capacity by quarters, actually third quarter, our capacity will go up by about 5%. That's mainly for the 28 nanometer, of course. That's the main reason for this 7% growth in revenue, but not as high as the bottom line gross margin. I believe that's temporary, and when we ramp the 28 to a bigger scale and with the improvement of profitability, I think this issue can be resolved.
Okay. I think we're about to wrap up for today's conference and conference call. Before we conclude, please be advised that the replay of the conference will be accessible three hours from now. Transcript will be available within 24 hours from now, both of which will 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.