聯華電子 (UMC) 2017 Q4 法說會逐字稿

Welcome, everyone, to UMC's 2017 Fourth Quarter Earnings Conference Call.

(Operator Instructions) And for your information, this conference call is now being broadcasted live over the Internet, and webcast replay will be available within an hour after the conference has finished. Please visit our website, www.umc.com, under the investor relation's Investors Events section.

And now I would like to introduce Mr. Michael Lin, Head of Investor Relations at UMC. Mr. Lin, you may begin.

Thank you, and welcome to the UMC's conference call for the fourth quarter of 2017.

I'm joined here by Mr. Jason Wang, the President of UMC; and Mr. Chitung Liu, the CFO of UMC. In a moment, we will hear our CFO present the fourth quarter financial results, followed by our president's key message to address UMC's forecast and the first quarter 2018 guidance. Once our President and the CFO complete their remarks, there will be a Q&A session.

UMC's quarterly financial reports are available at our website, www.umc.com, under the Investors Financials section.

During this conference, we will make forward-looking statements based on management's current expectation and beliefs. These forward-looking statements are subject to a number of risks and uncertainties that could cause actual results to differ materially, including the risks that may be beyond the company's control. For these risks, please refer to UMC's filing with the SEC in the U.S. and the ROC security authorities.

Now I would like to introduce UMC's CFO, Mr. Chitung Liu, to discuss our fourth quarter 2017 results.

Thank you, Michael.

I'd like to go through the Q4 '17 investor conference presentation material, which can be downloaded from our website.

Starting on Page 3. The fourth quarter of 2017 consolidated revenue was TWD 36.63 billion, with gross margin at 17.2%. Net income attributable to shareholders of the parent was TWD 1.77 billion, and earnings per ordinary shares were TWD 0.15. For loading in Q4 '17, capacity utilization rate was 90% compared to 96% in the previous quarter and 94% in the same quarter of previous year.

Please turn to Page 4. We can take a look at the income statement on a quarterly basis. Revenue declined around 2.8% to $36.6 billion. This is in line with our previous guidance. Gross margin stayed around -- at a similar level of 17.2% or TWD 6.3 billion. We see a 3.8% quarter-on-quarter decline in operating expenses and -- which lead to our operating income to reach $1.9 billion or 16.7% quarter-over-quarter growth. Due to -- third quarter usually is our high season for dividend income, and that's not the case for Q4, so our net nonoperating income in Q4 was negative of TWD 152 million. And net income, as a result, declined 51.6% to TWD 1.19 billion. Or the net income attributable to stockholder of the parent declined 49% quarter-over-quarter to TWD 1.77 billion. EPS is TWD 0.15.

For the full year, on Page 5, revenue grew in NT dollars about 1% year-on-year to TWD 149.2 billion. In U.S. dollar terms, however, considering the 6% NT dollar appreciation, our U.S. dollar-denominated revenue base actually grew over 7% on an annual basis. Gross margin, around 1 -- 18.1%, to $27 billion. This again is largely due to the NT dollar appreciation impact. For the -- net income attributable to stockholder of the parent grew 15.8% year-over-year to TWD 9.6 billion or equivalent of EPS TWD 0.79 per share.

On Page 6 is our simple balance sheet. Our cash and cash equivalent quickly climbed to TWD 81.6 billion at the end of Q4 last year, with total asset around TWD 394 billion.

EP -- ASP, on Page 7, in Q4 remained relatively flat, as our previous guidance. On Page 8 is our sales breakdown by location. Asia in Q4 represents 45% of revenue, and the U.S. represented another 43%. On Page 9, for the full year result, I think that's probably the first time that Asia is actually larger than North America in terms of sales breakdown, represent 47% and 43%, respectively, in 2017.

On Page 10. IDM still remained around 9% in Q4 and the same number for the whole year on Page 11.

On Page 12. Communication remains our largest segment, of -- 49% sales. And consumer is another 29%. For the full year, numbers-wise it's very similar in terms of the 3 major segments' breakdown.

For technology breakdown on 14, Page 14. 14 nanometer has grown to 2%. And 28 nanometers stayed around the same, around 15%. And for the full year breakdown on Page 15, 14 nanometer is now about 1%, and 28 nanometers to about 15%.

For the capacity, we continued to debottleneck our inch capacity given the strong demand, so there will be some meaningful increase in our 8N or the China inch wafer fab, along with some debottlenecking for the inch wafer fabs in Taiwan. And 12X, our new joint venture 12-inch wafer fab, we will also continue to see capacity increase. However, for Q1, there will be shorter working days, along with the annual maintenance, so capacity-wise for Q1, we'll show a minor decrease over Q4 of last year.

And 2018 foundry capital expenditure plan, so far, is around TWD 1.1 billion (sic) [USD 1.1 billion]. It's quite a meaningful decline compared to 2017. The actual result for 2017 CapEx was actually $1.44 billion instead of the previously budgeted $1.7 billion. Also, the distribution between 8 inch and 12 inch is quite different compared to the past experience. And for 2018, about 1/3 of the $1.1 billion budget, we'll spend on 8 inch. And 2/3 will be on 12 inch.

And that's the numbers -- the summary for UMC's results for Q4 2017. More details are available in the report, which can be downloaded or posted on our website.

I'll now turn the call over to our President, Mr. Wang.

Okay, thank you, Chitung. And good evening, everyone.

Here, let me update the fourth quarter operating result of UMC.

In the fourth quarter of 2017, UMC's foundry revenue was TWD 36.54 billion. During the quarter, our capacity utilization from legacy 8- and 12-inch technology continued to reflect the robust demand despite a decrease in 28 High-K contributions. The utilization rate of 90% led to an overall wafer shipment of 1.67 million 8-inch equivalents.

For the full year of 2017, UMC posted a 7% year-over-year revenue increase in U.S. dollars, as wafer shipment increased nearly 11% annually. As a result, our 2017 net income of the parent company grew nearly 16% year-over-year despite experiencing unfavorable movement of the NT dollar in the foreign exchange market.

Looking into first quarter 2018, we do anticipate our foundry business to remain relatively flat. And we are continuing our effort to capture new 28-nanometer business by working to secure new design opportunities, which will help rebuild our 28-nanometer momentums as these new product tape-outs are expected to enter production in the following months. In addition, we will leverage our manufacturing excellence to invest in the area with better ROI potentially (sic) [potential], including 12-inch mature technology as well as tool and equipment upgrades and an 8-inch facility. As such, our CapEx in 2018 budget will be approximately USD 1.1 billion. I believe our approach to optimizing the offering across 8- and 12-inch mature technology while moderating the pace of leading-edge expansion, will lead to a better financial performance that will preserve the best interests of our shareholder and employees.

Now that was our fourth quarter results. Let me go over the first quarter 2018 guidance. Wafer shipments will probably show an increase of 2% to 4%. And ASP in U.S. dollars is expected to decline 2%. Gross profit margin will be in the low-teens percentage range. And the capacity utilization rate will be approximately 90%. For foundry CapEx in 2018, as we mentioned earlier, it will be around USD 1.1 billion.

That concludes my comments. Thank you all for your attention, and now we are ready for questions.

(Operator Instructions) And our first question is coming from Randy Abrams, Crédit Suisse.

I wanted to ask the first question just about your -- if you could give an update on your plans for your China fab. And if you could recap for both this year and also for -- on the potential capacity in that fab, if you still plan that full fab for the logic or if you now intend to also divert that capacity to the memory business. So the necessary logic capacity may not reach the full 50k. You may split it with the memory business?

Okay. Randy, let me go over the fab 12X for China (inaudible). Our goal for the (inaudible)...

(inaudible).

Okay, so for the 12X in China, our Xiamen fab, our goal is (inaudible). So for the 2018 -- midyear 2018, we will reach about 17k per month, okay, (inaudible) of [11.5 inches]. By end of 2018, our goal is to reach about [22k] for that fab. And the maximum capacity for the [premium space] at this time is about 25k [versus 30]. So that's (inaudible) 12X...

Okay. If I could ask a follow-up. If the maximum now is 25k for logic, if you could discuss your strategy now on the advanced capacity. Once you reach that 25k, is it your view now to cap the investment, so more to stop the incremental expansion and more focused on cash flow, to reach a certain level of scale and then more move to an area where you're more optimizing mix but less focused on expansion and -- or I guess, the latter, if you see opportunity for more business or market share, would you have kind of a potential to look at additional fab options?

Well, Randy, as you know, our current direction for -- about CapEx expansion is following the ROIs expectations. So we're cautiously evaluating all the expansion options and the choices. So at this point, our plan is only -- into 2018 is only up to about 22k by end of the year. We haven't had any plans beyond that right now, okay? As far as the application goes, whether it's logic or memory goes into that fab, the memory is (inaudible) at this time. And we'll continue to focus on the logic (inaudible) and then -- and following the market dynamics. And we'll -- then we will expand our options, yes.

Okay. And if I could ask just on the guidance. If you could discuss both -- the shipments, it looks like, has a bit of relative strength, where you're growing -- normally, first quarter can be a down quarter, but if you could discuss the areas of strength by application or maybe strength and weaknesses by applications. And the flip side is, it looks like the gross margin on better sales is coming down. If you could talk about the factors on the gross margin affecting that. And I guess too we've seen some pressure from bare wafer pricing and some others, but if you could talk about your ability on pricing to pass some of that on; or as we go through the year efforts, to say, drive some improvement on gross margin.

Sure. From the Q1 application point, the computer remains to be the strongest one in terms of the Q1 outlook on a quarter-to-quarter basis. And both consumer and communications will be weaker. And I think that consumer is actually the weakest area. On the gross margin question, and there're a few reasons that -- some that you mentioned, but the major reason for us is really due to a lighter 28-nanometer utilization in Q1. That leads to weakening our sales mix. And besides the 28, the -- in Q1, we have a few working days that Chitung mentioned earlier and 2 maintenance. That affects the gross margin and the NT dollars as well. So the last is the -- what you said about the higher material costs. Especially in the 8- and 12-inch wafer substrates increased. And so that combination, that affects our gross margin in Q1, but I will still say the major contribution is really coming out from the lower 28-nanometer loading. Now as far as the passing-on the costs increase of the wafer substrate to the users of -- we -- actually, we start -- we had that discussion with our customers. Some is already happening and -- to mitigate some of the cost increase, but not on the old customer yet. And so for -- and so that is not 100% at this point, but we are deploying that. And we are having our conversation with the customer right now.

Okay. And the final question I had -- just since you mentioned the 28 lower utilization, if you could give expectations, say, for 28, where it may come down roughly as a percent of sales in Q1 or Q2? And then from the new applications, I think, where you're working on a, I guess, a second wave, by end of the year if maybe there's an early target or where it might be able to get back to.

Well, I'm probably not going to comment about a percentage-wise, but from the contribution, the revenue contribution in Q1 is definitely a softness than the Q4. And some of this is coming from the mobile sectors, particularly in Q1. We see a much weaker area in the mobile sector. And we do see that Q2 '18, the revenue contribution will improve, okay, on a quarter-to-quarter basis.

(Operator Instructions) And next we have Sebastian Hou from CLSA.

So my first one is on the 28-nanometer's outlook for 2018. I think one of your competitors talked about, recently talked about the lower addressable market for 28 nanometers in 2018 versus 2017. And I just wondered, what's the UMC view? And also, what's your outlook for this year -- year-over-year growth for 28 nanometers? Is it possible to increase year-over-year?

Well, we've been talking about this for a few quarters now. The -- we are under the 28 recovery mode. And the -- at this point, we do see some good momentum on our 28 HPC plus and with some of the customer engagements. And one data point is the number of the new tape-out at 28 is expected to grow double than last year. So there are some momentums. So the -- in general from a technology readiness standpoint, I think we are on track with our recovery. As for the revenue, our expectation is we're going to see something on the second half of 2018, as we planned it, from the HPC plus. And then we should continue to see an increase once the 22 nanometers is released, which probably is towards the end of this year and we expect that to improve in 2019. However, the business recovery is still subject to the success or the project ramp. And we do have high hopes there, but I think we're still going through that recovery mode. And this 28 business recovery momentum will probably continue throughout the year. And so your question about year-over-year from the 2017 to 2018 in terms of revenue contribution, I'm still cautious about that recovery, but we're just working diligence on that one right now.

Okay. Very clear. My second question is on the first quarter outlook on the margin, the profitability guidance. So with the low-teens gross profit margin, and also your OpEx-to-sales ratio seems to be also at the low-teens range, is it possible we see the almost breakeven of -- at the operating level in the first quarter?

Well, you're very accurate about that. That's also what we are seeing. Right now, based on the projection, this result may approach to operating breakeven. And however, I mean if we add back the losses carried by the minority interest, we will be in black. At this time, we'll continue to improve our -- the efforts in the area to improve this result, but I think we definitely will be in black. But we are approaching that breakeven point, yes.

Okay, I understand, but I think there was also a factor that -- which is the net -- the other operating income, which if I remember correctly that's the subsidy received in Xiamen fab. This number seems to be quite big in the fourth quarter, so how do we forecast that in -- throughout 2018 and also the first quarter?

So for first quarter, our President mentioned we are approximately breakeven in terms of the financial statement, but if we add back the minority interest -- the loss carried by minority interest, we should be in black on a pro forma concept. And as for the subsidies, the number should be similar to the quarterly average of 2017. I think that's also valid for the whole of fourth -- 4 quarters in 2018.

So even -- so for first quarter. And so even we assume -- well, we include the subsidy, which is at an operating level in the first quarter, is it still possible to reach the break-even level? Or actually, it should be better.

It won't be better. It's already included for the statement of approximately breakeven.

Okay, okay, I see, okay. And my last question is on the CapEx side. So the -- for 2017, the CapEx ended up lower than originally guided. I wonder, what was the reason? And also, can you give us the update on the CapEx distribution between 8 inch and 12 inch in 2017?

Well, 2017, I think roughly 9% is in 8 inch, and 91% in 12 inch. And the actual number was about $1.44 billion, as I mentioned, versus our budget of $1.7 billion. And the reason for the decrease was mainly due to a slower installation across-the-board. So some of the $1.7 billion -- or the $300 million also CapEx has been included in the $1.1 billion budget for 2018 already.

Okay. And to follow on that is that -- based on your 2018 outlook and also your guidance for the distribution. So it seems like you're 8 inch CapEx will double or more than double this year, so what's under the capacity addition we can expect on your 8 inch?

Well, for the 8 inches, we -- not much of a capacity increase because we have limited clean room space. Most of that is actually spending in upgrade in tools and the product mix. And so that will help us to improve our 8-inch overall product portfolios, yes.

And the next question is coming from Charlie Chan from Morgan Stanley.

So my first question is actually your exposure or your potential opportunity in so-called high-performance computing, especially now a lot of those AI semiconductors, right? So what kind of technology or process that UMC can supply to address this opportunity?

Well, at this point, I mean, most high-performance computing requires advanced technology. And the -- we are having 14 nanometers and -- but with a very limited capacity. So we don't have a whole lot of exposure there. We are engaging with several customers in the area, most through ASIC model. And because some of the customers in this space is -- they are not typical but [COT] type of customer, they sometimes need to have leverage in ASIC service. So we are engaging some of the customers through that. And so I would -- at this point, I will say our exposure in that space is relatively limited.

Yes. I think it is fair because your 14 nanometer just -- development is complete, right, but my question is really whether the recent development of this AI, HPC would change your [company decision] to sort of slow down those [immediate] developments because I remember your strategy is build up capacity based on demand, right? And now those [15 nanometer], 14-nanometer capacity seems to be quite under-supplied, right? So I'm asking whether you'll consider to invest more capacity in those 14-nanometer going forward.

Well, again, I mean, the -- our past strategy in CapEx spending is really the growth and the affordability. It means, as long as we can afford, we'll actually put in the capacity to meeting the customer needs and growth. And we have changed that from July last year. Our -- one of the considerations is by adding to that strategy is ROIs expectations. So if the ROI justify it and that will actually help us in financial performance, then that means, yes, we'll put in the capacity for that. And so that still remains to be our important factor to evaluate the CapEx expansion. Right now, for the 14, engagement is -- continues. And engineering activity is completed because, as we also reported midyear last year and -- we will wrap up on 14 technology. And we will try to take advantage of the 14 technology. That's why we'll continue engaging with the customer. And as far as the capacity expansion goes, we still have to go through that -- disciplines in terms of CapEx spending, yes, okay?

Yes. And my next question is regarding your [HPC side]. So I'm not sure if I interpreted this guidance right, but it seems like 8-inch wafer shipment should increase in 1Q because 28 nanometer declines. So can you give us some comments on what kind of 8-inch products are driving this kind of above seasonal demand?

Well, for our 8-inch of -- well, actually throughout 2017, our 8 inch demand has been very robust. And we've been running at full capacity throughout 2017. We continue to see strong demand in 8 inch for 2018. And it's coming from variable applications -- various applications. We see particularly strong in the mobile space for the RF switch; and the MCU, the embedded area; and -- as well as the display area. The most challenging thing today is actually managing customers because the demand is overwhelming right now, yes.

Next we will have Bill Lu from UBS for questions.

I got on the call a little bit late, so apologies if these questions have been asked. My first question is on depreciation. Can you talk about 2018 depreciation versus 2017? And then also, how do you think about that on a quarter-on-quarter basis?

For the 2018 depreciation, on a full annual basis, this is likely to show a 0 to 3% decline compared to 2017. And 2017 over 2016 was around 0 to 3% increase. So we start to show some minor decrease at least from a budgeting point of view, that 2018 will show a low single-digit decline if all the plan goes on track. And as for a quarterly breakdown, it's not -- it's very linear, so quarter-over-quarter should be very flat.

Okay, great. Second question is on 28 nanometers. Management has said in the past that we're going through a transition but that demand should start to rebound for 28 around the middle of 2018. I'm just wondering if you still feel that way.

Yes. Bill, yes. I mean we still feel that way. I actually reported earlier the -- about that when Randy asked the question, but then it's -- in case you didn't get that: So for the -- we will focus on the 28 HPC and HPC plus recovery. And our goal for 2018 is deliver the -- in the early part of 2018, we want to deliver the 28 HPC plus. And later this year, we want to deliver the 22, as we reported in the past couple of quarters. Right now for the HPC and HPC plus, we have demonstrated some good momentum in customer engagement, I mentioned earlier. And the index will base on the tape-outs. And for the tape-outs, right now we see -- project about -- that growth double from last year. So from a technology revenues point of view, we are on track. And we're working diligently on the 22 as well. So I -- at this point, the project is also on track for end-of-year release. And so our expectation is that we should see some recovery from second half of 2018, first from HPC and following 2019 with 22. Of course, those are the current engagements, but the actual business recovery was still subject to those -- to settle in that budget range. And we have high expectation out of that. More 28 tape-outs does translate to -- we are diversifying our 28s customer base, not as [bountiful] as what we have done in past couple years, but on the flip side, in the early wave of customers, they have a single part at higher volume. And now we have a smaller volume but multiple projects. So we just continue working on the -- on our plan, and right now the plans are on track, yes.

And if I can sneak in one last question. On 8 inch, you talked about CapEx not being spent on capacity but spent on upgrades. As you get through these upgrades, do you have any targets as far as where pricing and margins can get to for the 8-inch capacity?

Well, I mean one of the biggest reasons for driving the upgrade is try to improve our product mix. And by focusing on the PMICs and the RF switch, those areas -- that gives us better returns. But at the same time, we have to manage our current customer. So we can't just switch over one night, so the -- overnight. So that means we're going through this transition of improvements. So we'll continue upgrading our tool, focus on new engagement in the -- a better mix -- for a better mix, but at the same time, we also need to manage the current portfolios, yes.

Yes, maybe I can -- I fully understand that this will take time, but if you look at the PMIC and the auto-related customers, what is the -- how much better is the margin currently in those segments versus the overall 8-inch business?

Well, I won't be able to get into a specific percentage of each application, but the -- from the 8 inch ASP point of view, we hope -- our goal is at least we have to stay flat and not -- if not increase. So that's our target from the ASP point of view.

(Operator Instructions) And the next one is from Steven Pelayo from HSBC.

Yes. I'm curious if you can maybe talk a little bit about a -- kind of a target business model, I guess. With CapEx falling to this USD 1.1 billion, it looks like CapEx-to-sales falling to maybe low 20s or so, maybe 20%, depending on what revenue grows. Is this kind of a sustainable level you think for you? Is there an official kind of target for the company now to keep it around that level? How should we think about kind of capital intensity for UMC over the next few years?

We don't really have a set target for capital intensity per se. I guess the -- if you really go back to the corporate direction that's how the capacity spending will be ROI verified. Without these ROI incentives, there shouldn't be any expansion other than so-called maintenance CapEx. We, however, continue to look for low-cost capacity for potential acquisition if there're any good targets. And so there could be a different form of spending. Instead of buying new equipment, we may look for existing, older capacity. So it really goes back to the -- our enhancement. I'm not sure if Jason want to...

Yes. I think Chitung is right. I mean in principle it's the -- it's truly an ROI issue, right? I mean we'll -- we look at the -- in the past, we were definitely seeking for growth. And financially we can afford it, and so we pursue that CapEx decision. And at this point, if we only look at that again, then we're going to become vulnerable with our business model. And so our goal is try to strengthen our financial flexibility, as a first. And so reducing CapEx is just one of the tactics at this time and -- but the principle behind that is future CapEx expansion has to go through ROI justifications. And it's really not in -- the revenue intensity issue. It's about the -- how can we make our model cost competitive, yes.

Okay, understood. You also mentioned about the reason why you're cutting CapEx is to moderate the pace of leading-edge expansion. I'm curious. Will we see some of this also result in a moderating of R&D, for example? Will we see maybe impacts on the operating expense line as well over time?

Relatively speaking, the R&D costs are much smaller than the CapEx spending in the past. And so the -- we are cautious about our R&D spending. The goal, again, is for the ROI reason. If the project can definitely bring us potential returns, and then we will continue to fuel that. And so we truly don't have objective of cutting it or increasing it. It's more of a focus on each project's returns. Right now the -- from the R&D spending side, we haven't seen a significant decrease yet. So it's relatively flat at this point. We'll continue going through a project review to ensure those projects are worth spending our investments. So again, the -- we follow very -- we try to have a disciplined -- following our guidance of ROI consideration and whether it's on the CapEx side or on the R&D expansion side.

Okay, last question for me is just on 28 nanometer. It looks like the competitive landscape is certainly -- a lot more capacity is kind of ramping up in China and just a little bit everywhere actually. I'm curious about pricing pressures specifically at 28 nanometers. I remember, last year, that it was fairly intense through the year. Is that pressure still continuing given that it's not like 200 millimeters where you're struggling to meet all the demand?

Yes. I mean the pricing pressure is always there. We do see the China players being aggressive in that area. I mean -- but instead commenting about the other players, our focus is to try to make sure we are cost competitive. And so we follow the market price. And so in order to be competitive in that space, just like what we discussed earlier, our model, our business model needs to be competitive. So we have to be -- have a lower break-even point, so that means we can stay competitive. And that's why we deploy this ROI-driven approach because we want to quickly reduce our [BEP] as fast as we can. And for the 28 competitiveness point of view, I -- we still believe our technology is competitive. And we're still the -- probably the only one in the market today that are offering 28 HPC plus or HPC as well and HPC plus. And we are on track with our 22, so I think we're going to be competitive in terms of technology offering and compared to with -- in this market space. And the -- that's why we actually will continue investing in R&D. At the same time, we're seeking ways to reduce our costs and -- in order to stay competitive.

All right, if I can just sneak one more in. I was talking to a fabless chipmaker, and I was discussing kind of the strategy to moderate your leading-edge expansion. And the response was that, "Well, we still like to buy into people's roadmaps." So I'm wondering if you think that could potentially impact if you're moderating your pace of leading edge, that clients maybe won't be able to kind of migrate with you if you're not as aggressive on the leading edge. Does that ultimately impact demand?

Well, there was definitely some impact, but I think the impact is limited. Not every application will go into 7 or 5. So given our technology offering from 8 inch to 12 inch, all the way to 14 FinFET, we think we put -- we address most of that market on an application standpoint. So we may not address HPC, which is the highest-growth area, but if you look at the overall foundry market of a [$14 billion] size market, we have a good, bright -- we have a -- pretty much cover the 99.5% of that [$40 billion]. So I think, if we are only at around $4.5 billion to $5 billion in size, there is still plenty of room for us to serve our customer.

Yes.

Next we'll have Sebastian Hou from CLSA for questions.

Just a few -- follow-up on the -- do you have any revenue growth outlook for 2018?

(inaudible)?

Let me see. We definitely have a projection, but let me see how should I answer this. From an external market factor, the foundry growth for 2018, we are projecting about high single digits, okay? And it's driven by many different applications, including from cryptocurrency, AI, 5G, smartphone. So we are participating in that space, but for UMC and -- we still expect to grow year-over-year in 2018. However, it may not be in line with the foundry industry average. We're behind that and mainly because we're recalibrating ourselves right now. We still don't see some increase in the communication, computer area, but I think our growth will be moderate and, well, not in line with the foundry industry in general.

Okay. And I noticed that your 14-nanometer revenue has doubled quarter-on-quarter in 4Q. What was the main application? It's still cryptocurrency mining?

I don't think we ever commented about that before. So I mean I'm not denying and I'm not confirming that here. And -- but again, it's not because the application itself is -- it's really because we have a very limited customer there. And I -- by commenting that is kind of too specific, but from an application standpoint, the application process (inaudible). Our addressable area in this space is mainly in the cryptocurrency or AI area. So it -- yes. Well, I don't know if I complete -- if I should -- sorry.

No, no, no. That's clear already. And also Jason, you earlier mentioned about on the 8-inch side you see strengths in MCU, RF switch and also displayed area. So when you say displayed, is that driver IC?

Yes, yes. I mentioned about the driver IC, yes, right...

So overall -- sorry.

Go ahead. Sorry.

No, no, no. What I'm trying to just say is overall you see the driver IC is -- I mean on your side you're seeing a year-over-year growth for this year.

Yes. I mean, yes, we continue to see very strong demand coming from the display driver side, yes. Not on the -- not only on the existing side, also because the AMOLED is replacing some of the LCD panels. So we see some of the new application as well, right.

Okay, okay, got it. Last from me is that I noticed that in your 2017 the revenue from the Europe area also doubles year-over-year. In Japan, the revenue from Japan also grew strongly, like 20%, 30%. I just wonder, what was the driver? Is it mainly the IDM outsourcing? And also, what's the outlook for 2018 from these 2 regions?

Well, the -- for the Europe and Japan regions, it's mainly coming out from the mobile space as well. And most of our customers in Europe are IDM customers. And so I just don't want to comment specifically which customer, but it's on -- in the mobile and smartphone space, yes. And the -- we will continue engaging with the customer in Europe and Japan as well. And we hope that we can continue growing our Europe and Japan business, and we just continue engaging right now. And will that continue to grow? In 2018, it -- again our overall 2018 revenue is not going to grow significantly, so I expect the overall ratio [space of 10], yes.

Okay, so that's a little bit different from what I thought. I thought it was -- or it will be mostly driven by automotive and industrial. So not automotive?

No, not -- most of -- the majority of it is not, no.

And the next question is coming from Charlie Chan from Morgan Stanley.

Yes. So I actually have a similar question to what Sebastian has asked. So if it is not automotive, what kind of a product in mobile for consumer that will (inaudible) than mainly MCU? Can you elaborate a little bit?

Well, I think the question was about our Europe's revenue growth. And Europe's revenue growth is not MCU growth. It's the -- our 8-inch growth is mainly coming out from -- including MCU, but Europe is not, yes.

Okay. So for that MCU strength, what kind of end markets that you see the demand coming from?

We see that coming from IoT, coming from automotive and the -- communications, so they are different, yes, coming from different applications, yes.

Okay, okay. And also, my next question is about your free cash flow projection in -- for this year and coming years given you're reducing CapEx, right? So do we expect free cash flow to increase in coming years? And what was your kind of a preliminary dividend payout for this year?

Well, certainly we expect to see strong free cash flow in 2018 [to meet up the decline in] CapEx. And we will, of course, propose to have a higher dividend payout to our board, which will happen in March.

But you said that dividend payout -- dividend per share is going to be higher than last year EPS. Or are you still within this -- last year's earnings, not doing any special dividend?

There will be no special dividend, but we will certainly propose a higher dividend payout to the board.

Okay. And next one, I'm not sure if it's sensitive now to discuss about your DRAM projects. I remember there are some lawsuits between you and other memory companies. Can you explain a little bit? And at the end of the day, what's the progress there? When do you think the DRAM project will start a mass production?

Yes, there is a -- the lawsuit filed against us about DRAM, related to DRAM manufacturing. The -- our position is clear, and we have been reiterating this many times. And we have been always developing our DRAM internally. And future DRAM technology, we'll also build from ground up. And so our position right now is we are defending our intellectual property, including our patent in the memory technology to preserve this right in the best interest of our shareholders, okay, as well as our employees, right? I mean we're defending our fair right. And as far as the legal proceeding goes, we will fully cooperate with the authorities. And we have repeated that many times as well. And we stated we never copied any technology and have always conducted our R&D activity with integrity. So we are confident that, after due process, the court will rule in UMC's favor, but unfortunately there is the case against us, so we just have to deal with it. And we have [a case] in terms of our [delivered] and our own technology development integrity, so -- and we just have to go through this process, yes.

Okay, so before that clears out, can you continue your DRAM developments? Or does the lawsuit delay your schedule for DRAM production?

No. I mean, as I said, we have developed this internally and from the ground up, so the project is not being affected. We will continue our project. And -- I meant that the project is on track. And so we'll continue the -- working on the project without any delays, yes.

Ladies and gentlemen, we're running out of time, so we're taking the last question. And the last question is from Steven Pelayo from HSBC.

Yes, just 2 quick follow-ups. First of all, congratulations, after 8 years of negative free cash flow, getting back into the positive. I'm curious if -- what's the minimum cash needed on hand to kind of run your business? Will you be thinking about your dividend from that perspective? Companies like TI like to brag about paying out 100% of free cash flow. I'm curious if you think about it in terms of kind of cash on hand or you think about it in terms of just a payout of previous year's net income.

Well, I have to say the capital market in Taiwan doesn't really reward too much creativity in the special dividend side. We tried that before, so it's unlikely we will do that again. So a high dividend payout and -- coupled with our balance sheet enhancement or improvement. And maybe from time to time we will consider the share buyback program, but that will be pretty much it.

And Chitung, is there kind of a minimum cash that you think you need on your balance sheet?

It really depends. I -- as I mentioned, we are also looking for existing used fab by the -- can bring down our break-even point. So we do need to stay at a certain pocket in order to trigger those kinds of mergers if necessary. So it won't be too low, so it's unlikely to be extreme, to answer your question.

All right, my last question is just, I guess, under this new business model much more ROI focused, I mean, you're guiding the growth slower than the foundry market, but I'm curious, what does this mean for kind of structural margins as we go through 2018? I can't help but -- look, first quarter of 2017, your revenues were about 2% higher, but your gross margins were about 200 basis points higher. Now -- and maybe the Taiwan dollar was less of a headwind then, but can you talk a little about kind of your structural margins going forward? And if revenues were kind of growing, let's just say, I don't know, around 5% or something less than what you say foundry growth is, what kind of gross margin potential do you think you can drive with this company now?

Well, I mean these whole restructuring things take at least 1, if not 2 years. So in the near term or 1 to 2 years, we're still driven by the demand-supply, especially our recovery in the 28 nanometer. That will -- actually go along with the NT dollar currency movement will have a lot more impact than anything else. For the mid- to longer term, I think after 1 or 2 years, I think this lower CapEx, the better cash management, together with our approach to the specialty side of the technology altogether probably will start to kick in, but that's actually a 2-years-away scenario. In the near term, for the following 1 or 2 years, I will say the key element for UMC is still how to quickly recover our 28-nanometer customer base and also continue to improve our production efficiency.

So let me just summarize it real quick: Do you think it's possible to be over 20% gross margin in any quarter this year?

This question we can never answer.

Yes. I had to ask.

Thank you for all your questions. That concludes today's Q&A session. I'll return it over to UMC Head of IR for closing remarks.

Thank you for joining us today. We appreciate your questions. As always, if you have any additional follow-up questions, please feel free to contact UMC at ir@umc.com.

Have a good day. Thank you.

Thank you, ladies and gentlemen. That concludes our conference for fourth quarter 2017, and thank you for your participation in UMC's conference.

There will be a webcast replay within an hour. Please visit www.umc.com, under the Investors Events section.

You may now disconnect. Goodbye.