Welcome everyone to UMC's 2019 fourth quarter earnings conference call. All lines have been placed on mute to prevent background noise. After the presentation, there will be a question and answer session. Please follow the instructions given at that time if you would like to ask a question. For your information, this conference call is now being broadcast live over the internet. Webcast replay will be available within an hour after the conference is finished. Please visit our website, www.umc.com, under the Investor Relations, Investors Events section. Now I would like to introduce Mr. Michael Lin, Head of Investor Relations at UMC. Mr. Lin, please begin.
Thank you, and welcome to UMC's conference call for the fourth quarter of 2019. I am joined by Mr. Jason Wang, the President of UMC, and Mr. Chi-Tung 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 focus and the first quarter 2020 guidance. Once our President and 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 Financial section. During this conference, we will make forward-looking statements based on the 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 company's control.
For this risk, please refer to UMC's filing with the SEC in the U.S. and the overseas security authorities. Now, I would like to introduce UMC CFO, Mr. Chi-Tung Liu, to discuss our fourth quarter 2019 financial results.
Thank you, Michael. I would like to go through the 4Q19 investor conference presentation material, which can be downloaded from our website. Starting on page 3, the fourth quarter of 2019, consolidated revenue was TWD 41.85 billion, with a gross margin at 16.7%. The net income attributable to the stockholder of the parent was TWD 3.84 billion, and earnings per ordinary shares were TWD 0.33. Our loading utilization rate in Q4 of 2019 was 92%, slightly better than 91% in the previous quarter, and also 88% in the same quarter of last year. Revenue, TWD 41.8 billion, was a record high for UMC. On page four, financial income statement on a sequential basis. Revenue 41.8 grew by 10.9% quarter-over-quarter. Gross margin, as a result, also grew 8.3% sequentially to TWD 6.96 billion, or 16.7% gross margin.
Because the combination of our newly acquired U.S., Japan operations, our operating expenses grew to TWD 6.1 billion, which led to operating income of TWD 2.018 billion. With the better stock market performance as well as the strengthened RMB versus U.S. dollars, our non-operating income was again of TWD 946 million. The total net income attributable to the stockholder of the parent was TWD 3.83 or an EPS of TWD 0.33. On page five, our year-over-year comparison. The full year revenue was TWD 148 billion, a small decline of 2% year-over-year, which mainly due to the weakness in the beginning of 2019. Gross margin rate is around 14.4% or a total gross margin of TWD 21.3 billion. Our operating expenses is under control and reduced to TWD 21.8 billion for the full year.
As a result, the total net income attributable to the stockholder of the parent for the full 2019 was TWD 9.7 billion or equivalent to an EPS of TWD 0.82. For balance sheet for the full year, our cash has accumulated to TWD 95.4 billion, and the total equity for the company is at TWD 207 billion. On page seven, our ASP for the quarter was somewhat ravaged compared to the previous quarter. On page eight, our revenue breakdown. Because of the combination of the new acquisition in Japan, our total revenue from Japan grew to 9% of the total pie compared to 2% in the previous quarter. Asia remained as big as 55%, our single largest market, and North America is around 30%. For the full year, Asia is about 57% and North America is 32%.
Japan and Europe are about 5%-6% each. For IDM, again, because of the combination of Japan operation, in Q4, IDM percentage jumped to 13% of the pie and fabless is the remaining 87%. For the full year, the difference is much less. It's similar to the previous year, around 8%-9% for IDM. On page 12, our segment breakdown. Still, we see similar distributions among communication, computer, and consumers, with communication as the single largest segment around 54%. For the whole year, we'll continue to see more communication contribution to 52% for the whole year, and consumer is around 26%. On page 14, our total revenue coming from 40 nanometer and below is around 32%, with 28 nanometer is around 10%.
For the full year, we have minimum contribution from 14 and a stable contribution from 28 and 40 nanometers, which in combined is around 34% for the full year of 2019. On page 16, our Q1 capacity here is factoring the annual maintenance already, but still, we see a somewhat flattish total capacity available in Q1 versus the previous quarter. Our full year CapEx for the 2020, right now, the budget is around $1 billion, and 85% will be 12 in related. The above is a summary of UMC results for 2019. More details are available in the report, which has been posted on our website. I will now turn the call over to President of UMC, Mr. Wang.
Thank you, Chi-Tung. Good evening, everyone. Here, I would like to update the fourth quarter operating result of UMC. During the fourth quarter, we started to account for the foundry operation at our recently acquired USJC, which is Fab 12M in Japan. In spite of the currency headwinds encountered in the foreign exchange market, our foundry revenue increased 10.9% quarter-over-quarter to NT$41.83 billion, leading to a foundry operating margin of 4.9%. Utilization rate increased to 92%, bringing wafer shipment to 2.04 million 8 in equivalent wafers, primarily driven by the communication and computing segments. For the year, our earning per share increased 41% year-over-year to NT$0.82. Our continued disciplined CapEx approach also enabled UMC to generate a total of NT$37.1 billion in free cash flow, up 19% year-over-year.
In terms of technology, we recently validated our 22 nanometer process on a USB2.0 test vehicle, demonstrating the technology readiness of this design rule shrink from 28 nanometer. UMC's 22 nanometer process features a 10% area reduction, better power to performance ratio, and enhanced RF capabilities compared to our existing 28nm High-K metal gate technology. Looking to the first quarter of 2020, based on the customer forecast, the overall business outlook appears to remain consistent with the previous quarter, primarily due to a stable wafer demand across wireless communication and computer peripheral segments. As we receive new product tape-outs that will enter future production pipelines, we expect to benefit from 5G and IoT trends that will generate additional semiconductor demand, specifically in wireless devices as well as power management applications.
While our focus remains to maintain the discipline, the capital expenditure spending owing to meet long-term customer and market demand, we have set up our 2020 CapEx budget of $1 billion. UMC will continue to penetrate into new segments and expand our presence in existing markets. Our core competence in process technology development and world-class foundry service will strengthen our position in delivering logic and specialty manufacturing solutions. Now, I would like to address the Wuhan coronavirus situation here. Regarding the novel coronavirus outbreak, we promptly set up a cross-functional prevention team prior to Chinese New Year. That is specifically tasked with managing the potential risk of the coronavirus at UMC. Today, all of our operations remain normal, and we will continue to actively monitor the radar both internally and externally so we may better timely respond to any changes in this fluid situation.
Now let's move on to the first quarter 2020 guidance. Our wafer shipment will remain flat. ASP in U.S. dollar is expected to remain flat. Gross profit margin will be in the mid-teens percentage range. Capacity utilization rate will be around 90%. Our 2020 CapEx budget will be US $1 billion. That concludes my comments. Thank you all for your attention. Now we are ready for questions.
Thank you, President Wang. Ladies and gentlemen, we will now begin our question and answer session. If you have a question for any of today's speakers, please press zero one on your telephone keypad, and you will enter the queue. After you are announced, please ask your question. If you find that your question has been answered before it is your turn to speak, please press zero two to cancel the question. Thank you. The first question is coming from Randy Abrams of Credit Suisse. Go ahead, please.
Okay. Yes. Thank you. Good evening. I wanted to ask the first question, if you could give a look ahead to a view for 2020, maybe an outlook for foundry sector, whether it is for the whole industry or stripping out the advanced technology you do not access. Then your view relative to that, how you expect UMC to perform. The second part, if you have discerned any changes from the virus at this stage, whether from the customer demand side or some of the activity from that at this stage.
Sure. Hi, Randy. Thank you for the question. The first question is about our view on the foundry outlook in 2020. We expect semi to grow in mid-single digit and the foundry segment to grow in high single digit percentage range. That is actually including the advanced node as well. Okay. That does not exclude the advanced node. That is now focused on addressable only. If we look at the UMC growth in 2020, with USJC, our newly acquired fab, we will grow higher than the foundry industry. Without the USJC, we will grow in line or slightly better than the foundry industry. So that is where we project at this point for the 2020. As far as for the question of the coronavirus impact, at this point, we are closely communicating with our customer to monitor the development of the situation.
At this moment, there is no change in our Q1 demand. Should the outbreak deteriorate into a situation that may cause supply chain issues, which we are looking at it closely in China, that could be some impact to the entire semiconductor industry. At this point, there is no change yet.
Okay. No, thank you for that. If I could ask you on the 28 nanometer, where last quarter, I think you saw some light at the end of the tunnel or some improvement. Could you maybe give an update on the applications you're qualifying, and then timing to bring that in? Maybe in that, how you could see 28 ramp up as percent of revenue through the year or toward year-end. Also add a question, just as you ramp back 28, how that would impact the gross margin where you get the benefit of filling the fab, but I guess relative to corporate average, how you see it, whether it would be accretive to corporate gross margin or still dilutive to the margin.
Sure. Well, we've been reporting many quarters about the 28 contribution to us, and we've been talking about this fragment in the market, and we're still at this recovery mode. So, we expect to see the 28 nanometer contribution in Q1 2020 will still be under the recovery mode. But we are confident in improving our 28 utilization in second half 2020. What we foresee is our key 28 projects engagement will start to enter volume production in third quarter 2020. Mainly driven by the wireless communication ICs. As far as for the gross margin.
Yeah. For the gross margin, I think any recovery in the currently underutilized 28 nanometer capacity will be an incremental positive to our overall corporate gross margin.
Okay. Great. Then last question, just on the CapEx of TWD 1 billion with the 85% 12 in and then the rest 8 in. Could you maybe lay out in capacity addition to where the fabs are where you expect to increase the actual wafer capacity?
Sure. The way that you break it down first is between 8 in and 12 in. The other way we look at it is more we split into three different categories. One third of the CapEx will fall the non-capacity related. For example, our general budget purpose. One third will count for the increase of capacity, major for 28 nanometer in our 12X, the Xiamen facility. The other one third is to upgrade throughout our overall capacity pool. The overall capacity growth rate will be a low single digits year-over-year in 2021. What we have been budgeting for 2020 is more of a capacity plan for 2021.
Okay. On the 28, if you could give an update how much capacity you have. I think maybe last update it was, if I have it right, 39 and a half thousand and 5K in Xiamen. If you could maybe give an update how much you have for 28 and how much you may expand 28.
Well, the 28 monthly capacity will remain unchanged at the 45,000-46,000 a month in 2020. What we budgeted for the 2020 CapEx is mainly for the increase of 2021. Yeah.
So, just-
Oh, okay. I understand.
Adding to that, the Xiamen capacity likely to reach full of 25,000 wafers per month by mid of 2021.
Okay, great. Thanks a lot, guys.
Sure. Thank you.
The next question is coming from Charlie Chan of Morgan Stanley. Go ahead, please.
Hi, happy New Year. First of all, I want to follow up the question regarding the 28 nanometer demand. You said that demand comes on the wireless semiconductor. My question is that, what kind of wireless, because most of the smartphone AP already migrate to 12 nanometer or below. Secondly, do you classify the AMOLED driver IC and the image signal processor as the wireless semiconductor for 28 nanometer? Thanks.
Sure. First, Happy New Year to you, too. Yes. The answer is yes. We are considering the Wi-Fi devices and the display drivers IC and the sound sensor device. They are considered as the application for the 28 nanometers, yes.
Okay. Thanks for clarification. Second thing is that, we are being seen that lot of trailing edge are in Taiwanese and some older overflow from those bigger players. Can you please comment, which process nodes you see are in Taiwanese or even in shortage? Second thing is that, do you see any possibility that you have to select customers, even hike the price in the coming quarters?
Okay. Well, that's more a question for across all nodes. Let me see if I can break it down this way. We see tightness, the capacity tightness across the 8 in advanced node, okay, which is anything smaller than the 0.18 micron.
We see the 12 in, all the legacy nodes from the 90 nanometer all the way to 35 nanometer. Okay.
The 40 nanometers, we see some decline in Q4 2019, and largely due to the weakened demand. But in the long-term view, we are continuing caution monitoring the node migration and about this particular node, the 40 nanometers. That's kind of how we look at the market right now.
Okay. Do you expect that tightening will continue and you have to kind of be more firm on your pricing or even doing some price hike later quarters?
Well, we do think this tightness will continue.
On those nodes I highlighted. In terms of the pricing, we still have to align with the market price. We will continue working closer with our customers based on the market price trend. Yeah.
Okay. Lastly, maybe some updates on the lawsuit with Micron. Maybe Chi-Tung or Jason can help us, because this has been a very big overhand to the stock, and it has been for more than a year, right? So we really hope to get some clarification regarding the status, the timing, potential impact to company's finance and operation. It would be very helpful for long-term investors. Thanks.
Yes. We understand that. Right now, the litigation is still ongoing. Given the case, it is under the judicial process. We do not have any updated information. So if there is any significant development, we will disclose them accordingly.
Okay
truly not much of update right now.
Okay. Just to clarify, in your open remark, you Sorry, the answer to Randy's question. You said that your 2020 revenue is going to outgrow the foundry industry without including the Japan fab. Is that the right takeaway?
I said without the Japan fab, we will grow in line or slightly better than the foundry industry, yes.
Okay. What is the reason that you are outgrowing or gaining market share? Because I can understand TSMC can claim that because they continue to invest in advanced nodes. But for you guys, you are also competing with the Chinese vendors, but Chinese vendors seems to be very competent. They are well-positioned for the Chinese customers' growth. Can you give us some elaboration where you can outgrow?
Sure. First of all, we expect to capture high growth from the wireless communication, including early 5G rollout opportunities in both logic and specialty technology.
And obviously, the change we believe is mainly driven by the demand pickup. That's one reason. The second is, of course, the market share gain in 2020.
Mm-hmm. Okay. So in which segment you are gaining share?
Let me go this way. If we look at the market trend, we already start observing some of the evidence by, one, increasing power and RF application, such as a transceiver and image in a 5G phone and a base station. That's one area.
Second is increase of OLED adoption in a smartphone.
Okay. That's second.
Which will fuel the OLED driver IC.
We also touched that earlier, too, the driver. The third is the multi-camera modules that are found in a smartphone that will drive the sensor and the controller wafer demand. Also, we saw the upgrades in the Wi-Fi standard for the faster connectivity requirements, and we will get benefits on that as well. This is probably what I can think of it from four different areas. Yeah.
Yeah. But for that outsourcing of AMOLED driver IC and the ISP from Korea customer, do you think there is something structural? Maybe they have some internal fab shortage this year, but long term, how solid and how sustainable do you think this outsourcing is going to be?
Well, we usually don't comment on any specific customer. I don't think it's appropriate for me to comment that. Yeah.
Okay. Sure. Yeah. Just if you can give us some perspective from high level strategy perspective. Why this could be structural? I think that should be helpful. Maybe not for a specific customer, but just for the sector or industry overall. Is that possible?
Well, I think the changing in the market space does fuel the increase of wafer semiconductor demand, right?
Mm-hmm. Yep.
That's significant.
Okay.
I would say that's the opportunity that we are participating right now, yes.
Okay. This is very helpful. Thank you.
Sure. Thank you.
The next question is coming from Gokul Hariharan of JPMorgan. Please ask your question.
Sure. Thanks for taking my questions. First of all, could you talk a little bit about how you expect depreciation to shape up in 2020? With the Fujitsu fab consolidation and higher CapEx compared to last year, what is our expectation of depreciation going into 2021? I think previously we had expected, I think 2021 probably could be the year where we will see a meaningful drop-off in depreciation expenses. Is there a change to that given the increase in CapEx and expansion on 28 nanometer? That is my first question, and I had a couple of follow-ups as well.
For depreciation in 2019, it declined about 5%-6%, as we expected. Because of the combination of Japanese fab, our 2020 expectation in terms of depreciation decline is around low single digits. Of course, yes, as we mentioned earlier, the peak of the 28 spending was about four years ago. So 2021 or 2022, we will expect to see even much bigger rate of decline in terms of our depreciation expenses.
Could we be a little bit more specific, given we are spending a little bit more CapEx? Are we thinking about 15%-20% decline in depreciation in 2021, or is it going to be more gradual than that?
I think it probably will be a little bit more gradual than that. Still, our spending for 2018 and 2019 both are around $0.7 billion only. Still, I think the mega trend won't change much. But the magnitude for 2021 probably will be less than what you mentioned.
Got it. Thank you. Also, given the roughly 10% accretion on revenue side from the Fujitsu acquisition, could you also comment about what is the margin impact from consolidating this fab? Is it additive impact to gross margin, or is it similar to corporate average?
For the Q4, the first quarter, they are slightly above break even. As we mentioned earlier, we are driving the synergy, and it may take up to two to three quarters. We do expect the contribution to improve over time. That's the current situation. They are in black in Q4 already, and we expect they will continue to be profitable in Q1 of this year.
Okay. Thank you. Last question from me. Could you talk a little bit about how the current situation on 8 in utilization, demand, capacity, and pricing? If the coronavirus situation deteriorates and the current weakness that you are seeing in China 5G demand persists, if there is a lower expectation on 5G demand itself, does that affect your 8 in utilization or demand, or 8 in, you would say, is relatively less impacted?
Well, our Q4 2019, the 8 in is at the low 90% utilization rate. In Q1 2020, our utilization will be in mid-90 range. We are expecting to see higher 8 in utilization rate after the Q1 2020. So we remain confident about this outlook of the 8 in. In terms of the virus outbreak, as we reported earlier, based on the communication with our customer, at this point, there is no change in Q1 yet.
Okay. Got it. Thank you.
Sure.
The next one is coming from Zee Hong of China Renaissance Securities. Go ahead, please.
Oh, hi. Good evening, gentlemen. With the company now turning a lot more cash generative, I am wondering whether the company would consider to pay a special dividend or to raise the payout ratio over 100%.
Well, our last year payout ratio was 100% already, so it is probably difficult to get higher than that. For this coming year, of course, we will continue with our high dividend payout policy, and we will propose that to our board in due courses.
Okay, great. The other question, again, on the Japanese fab. I believe it is still not fully loaded, right? Would the company consider to move more business from Taiwan to Japan, or how easy would that be?
Well, that is part of our synergy plans. In addition to streamlining the material procurement and fab productivity enhancement, we also try to bridge in product from the headquarter to the Japan fab. So that is part of our activities, yes.
The customers are actually quite receptive to that arrangement, to move from the Taiwan fab to Japan fab.
Well, yes. We are aligned with our customers for those activities, yes.
Okay. All right. Thank you very much, and Happy New Year.
Yeah, you too.
Thank you.
The next question is coming from Bruce Lu of Goldman Sachs. Go ahead, please.
Hi. This is Bruce. I need to clarify one thing, that 28 nanometer profitability was similar to corporate average, even with full capacity. Is that right?
No, that is not the way we look at this. Any recovery in the 28 nanometer capacity utilization rate, which we expect to see that starting from mid of this year, will be incrementally positive to our current corporate average profitability.
Okay. That would be still around 23% range. Is that right?
No, I didn't say that. It's quite complicated. Even for the same 28 nanometer technology, different customers will have different profit margins, not to mention the different fabs will have also-
I see.
very different profit structures.
Okay. My question is that, I'm actually very surprised to see the 28 nanometer capacity expansion. If you use $1 billion CapEx, 70%-80% for the advance or for 28 nanometer, your 28 nanometer capacity might increase by 20%-30% in 2021. TSMC, the biggest 28 nanometer capacity provider, suggests that the industry is still oversupplied. Do you expect this incremental capacity for your 28 will be margin accretive, ROE accretive?
If you put the whole actions in one picture, yes, because our current Xiamen fab, it will be the main area for 28 nanometer expansion. Right now, Xiamen fab is only about 70% equipped, and the remaining 30%, 35%, once we reach the full capacity, the whole profitability structure will be a lot more ideal. It will be also much easier to reach a break-even position with a fully equipped 25,000 wafers per month fab, compared to the current status.
Yeah, but my question is that, by the end of 2020, assuming you are increasing your capacity for 28, that's based on that you will have the full capacity for 28 by the end of 2020, right? So you will have a better profitability by then. With this incremental capacity, will that be a margin and ROE accredited?
It won't be equivalent to our current corporate average, but it will be much better than its current single fab profitability. So we will see a pretty significant profitability improvement because of the economy of scale benefit for that one particular fab.
I see. What is the revenue concentration for your 28 nanometer to a single customer or the single end customers? Do you see the high revenue concentration on that? Because we are trying to evaluate the risk for that.
Yeah, let me answer you in a different way first. Our top 10 customer are around 55%-60% as always, and top two, top three customers sometimes is around 10% threshold.
For 28 nanometers, we are now at more than a dozen of customers. The diversity may not be as diversified as the whole company, but also are quite diversified. I don't have a 28 nanometer customer list for you right now.
The concentration won't be too much higher than our corporate average.
I see. Understand. Last question is that the management guidance are guided by foundry industry growth is high single digit, and this is much slower than what TSMC guided for the foundry industry. Can I know the differences?
Well, we all have a different view and look at this market outlook. So, it's our intelligence telling us the outlook for 2020 will be somewhere at the high single digits. So that's why we have our outlooks, yeah.
The gap is quite big.
I can't comment. I don't know how they calculate.
Yeah, we can only speak for ourselves.
Okay. Understand that. Because we need to. I just try to know the differences, though.
Of course. Yeah. Because we don't have a detail of how they calculate it, so it's hard.
I see.
to compare the numbers.
But if they are having 50% market share for the whole market, and they are guiding for 20% for their revenue growth. If you are guiding for high single digit for the industry growth, which means the rest of the founder will be low single digit decline.
You cannot put the two sides of data and making one story. I think it will be difficult for us to answer that.
But when you calculate your stuff, you already factor in that TSMC will grow by 20% as they guided, right?
We use our market intelligence to come up with this high single-digit number.
Understand. Thank you.
Thank you.
As a reminder, please press zero one on your keypad if you would like to ask the question. Thank you. The next question is coming from Sebastian Hou of CLSA. Go ahead, please.
Great, thank you. The first question is on follow on what Chi-Tung just said about the significant improvement possibility in your Xiamen fab. Would that change the subsidy from the government you are receiving?
No, the subsidies we have already received, in cash in our pocket. But accounting-wise, we have to recognize that along with our depreciation timetable, which is around six years.
Okay.
We already recognized that for more than two, three years. The remaining three to four years is intact.
Okay. That is still coming through the other incomes in operating line.
Yeah. It's our, you know, financial.
Okay. Basically, it's irrelevant to the actual profitability of your Xiamen fab right now.
Yeah, the tax might be different, but the subsidies wise, the whole part figure is about the same.
Okay. But the net is still positive to the company.
Yeah, sure.
profit. Okay, got it. In the prepared remarks, the co-president mentioned that the company is receiving new product tape-outs to enter future production pipeline. Can you elaborate more about what type of the tape-out you are receiving and that make you excited?
Well, the outlook of the tape-out is across different various nodes. For our 28 nanometer projects engagement, we talked about that earlier. Mainly driven by the wireless communication ICs, including Wi-Fi, display drivers, and others. For the 8 in, we also touched it, talked about the RF switch and the transceiver and those areas. They are for different various nodes. Mainly, I would say, if you look at this in general, it is really a lot of it is associated with the 5G rollout. We see this 5G rollout does help in the overall semiconductor demand increase. We see quite a bit of opportunity in there, yeah.
Okay, got it. When you mentioned about the four areas that you are seeing growth drive the likely above industry average growth earlier, and one of that is 5G smartphone and base station. Can you give us more details about what type of the ICs that you are supplying into 5G smartphone and base station?
Under the 5G infrastructure, such like the Sub-6 GHz, the RF SOI, the RF switch, were associated with that. Power management also associated with that, and that's for the 8 in. Some of the display requirement actually for the 5G smartphone, the AMOLED, and that's the display drivers, that's related to 5G as well. That's where the area we see many of the opportunities.
Okay. Got it. I have another question, it is about your plan on the J2 fab that you just acquired. I understand that you don't mention about it is already above break even now and probably will improve. As you evaluate that, first, how long does it take for that fab to reach the corporate average utilization rate? Second is that once you reach the corporate average utilization rate, then what's the profitability like? Will it also be similar to the corporate average or still be interestingly low?
Well, again, our corporate average hopefully is improving as a whole. It is difficult to just benchmark to corporate average. The goal is really to drive the synergy out of this newly acquired fab. So far, it has been meeting or even exceeding our expectations. We're pretty happy with the current achievement and we do expect to see further benefit out of the whole integration effort. Again, we cannot quote in terms of numbers as we see no reason why this fab cannot have similar profitability as any single fab we have in Taiwan or Singapore.
Okay. I see. Last questions from me. I think this probably may addressed by the other analysts, but I sort of dial in later. Just to follow on the depreciation outlook for 2021. With the new CapEx on 2020 now in place, what's your updated view on the depreciation decline magnitude for 2021?
Well, it's still on the decline trend. Back in 2019, the overall depreciation expenses declined by about 5% to 6%. For 2020, we expect to see low single digit decline, with a little bit more decline rate in 2021.
Okay. A little bit more than low single digit.
It will be a lot more than low single digit, yes.
A lot more. Okay.
Yeah.
Okay. Got it. Thank you.
The next question is coming from Stephen Chin of Elisha Capital. Go ahead, please.
Hi. Thank you for taking my question. This is Stephen from Elisha Capital. The first question I would still like to follow up on the CapEx. In the past few years, the CapEx has pretty disciplined, and this year we see it higher. You already explained the reason. I'm just wondering if from the company management point of view, you still aim for a disciplined CapEx and high generation business model, or we should treat this year as kind of an inflection point that you start to see more semiconductor content weight opportunity, so you start to aim for more growth?
Well, our CapEx policy remains the same. It is disciplined CapEx policy. We'll continue cautiously proceed with this approach. Our budget will be subject to the strengthen ROI criteria every year. When we do that, we actually factor in our cash dividends payout to shareholder while ensuring affordable spending that will deliver our target organic growth. It is a balanced act and we continue with our disciplined CapEx approach.
Yeah, understood. Thank you very much. My second question is regarding your customer portfolio change regarding the, I mean, based on the geography breakdown. In 2019, we see Asian customer percentage-wise increase quite significantly. Just wonder, do you consider this trend will continue?
I think in the past couple years, we actually observed this trend a couple years ago, it is not recent. We think we will get to a situation where it starts saturating, and this will be probably, the order will be similar to what we are seeing today.
Understood. Yeah. The next question is regarding, although a follow-up on your estimate for the foundry revenue growth. Just curious, do you consider those internal supply also as your number in your calculation, or you only consider the third-party foundry?
Can you repeat that question again?
Yeah. The question is, when you guide for the foundry market growth, I'm just wondering if you consider those, they call it foundry, but most of the revenue is for the internal demand. Do you consider that as part of your calculation?
No. I don't think we include that, no.
I see. Okay. The last question is, you mentioned about as a 200mm, is part of that, the growth is on the RF switch, and you also mentioned about RF SOI. I'm curious if you also see the SOI growth at 300mm. For UMC particular, do you see it is mainly on the RF SOI or you are also doing the FDSOI? Thank you.
We are not participating in FDSOI today, and we're only referring to the RF SOI. As the product migration trend, we do see the RF SOI will migrating from 8 in to 12 in. As far as the 8 in demand outlook, beside the RF switch, RF transceiver, we also see area, and we see a significant power management ICs demand as well.
Understood. Thank you very much. Also, congratulations to your good Q4 outlook. Thank you.
Thank you.
The next question is coming from Randy Abrams of Credit Suisse. Go ahead, please.
Okay. Yes, thank you. I just had a few quick follow-up questions. One for the OpEx, where it has been running about 10%-11% of sales. If you could talk about how you kind of see that growing in the coming year, just an absolute or as a percent of revenue. The other one, just the tax rate, to use.
For OpEx in absolute TWD dollar term, because of the addition of the Japanese operation, we do not see too much fluctuations in the absolute OpEx TWD dollars. Because of the kind of optimistic outlook for the whole 2020, we do expect the OpEx as a percentage of revenue will gradually trend down. That's our goal. For tax, there's not much change in the current circumstance. We do have some tax expenses reversal in Q4 of last year, which result in a tax credit instead of tax expenses for Q4 last year. Overall, we are talking about 10% plus or minus of corporate tax average.
Okay. The last question on the end markets, the stable first quarter, even with the Chinese New Year, a pretty decent for shipments. Could you give maybe a split by the applications, what you're seeing maybe areas that may be growing a bit in first quarter and any areas declining a bit?
Yeah. We see the Q1 2020, the strongest is in the computer segment. Then followed by the communication, and the consumer is still the weakest right now.
Okay. Computer segment, I guess within that, because overall PC, there's I guess not that much happening, but is there something driving that to be better?
Yeah. We actually see lots of high-speed IO interface devices. The USB, the HDMI, all those high-speed devices, we see quite a bit demand increase. Yeah.
Okay, great. Thank you.
Sure.
Ladies and gentlemen, we thank you for all your questions. That concludes today's Q&A session. I'll turn things over to UMC Head of IR for closing remarks.
Thank you for attending this conference 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 Q4 2019. We 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 event section. You may now disconnect. Goodbye.