聯華電子 (UMC) 2018 Q4 法說會逐字稿

Welcome everyone to UMC's 2018 Fourth Quarter Earnings Conference Call.

(Operator Instructions).

For your information, this conference call is now being broadcasted 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.

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 2018.

I'm joined 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 of 2019 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 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 the company's control.

For these risks, please refer to UMC's filings with the SEC in the U.S. and the ROC securities authorities.

Now I would like to introduce UMC's CFO, Mr. Chitung Liu, to discuss our fourth quarter 2018 financial results.

Thank you, Michael.

I would like to go through the Q4 '18 investor conference presentation material which can be downloaded from our website.

Starting on Page 3, the fourth quarter of 2018, consolidated revenue was TWD 35.52 billion.

We saw gross margin at 13%.

The net loss attributable to the stockholder of the parent was TWD 1.71 billion and loss per ordinary shares were TWD 0.14.

The capacity utilization rate in fourth quarter of 2018 was 88%, down 6 percentage points from 94% in the previous quarter.

And on Page 4, our quarterly revenue, TWD 35.5 billion, was down 9.8% quarter-over-quarter.

About 5.2% was coming from a decline in wafer shipments.

The rest is coming from a product mix change and cost pricing decline.

And gross margin, as I mentioned earlier, was 13% or TWD 4.6 billion.

Operating expenses went up 12.2% quarter-over-quarter, mainly due to a onetime book of our DRAM R&D related project.

About TWD 700 million plus onetime expenses was booked in the Q4 of 2018.

And in terms of non-operating income loss of TWD 1.99 billion, or close to TWD 2 billion, was mainly due to the valuation of our stock holdings in the capital market, given the volatility of the stock market around the globe.

And net income is a loss of TWD 3 billion, and net income attributable to stockholder of the parent was TWD 1.7 billion or an EPS of minus TWD 0.14.

For the whole year, our revenue grew 1.3% year-over-year to TWD 151.2 billion.

Gross margin was down 15.6% to TWD 22.8 billion.

Operating expenses would have been a lower number if we did not include the onetime charge I mentioned earlier in Q4 which related to our DRAM R&D project.

And overall operating income in 2018 was TWD 5.79 billion or 3.8% operating margin.

Net non-operating income, which mainly because of the valuation of our stock holdings as well as some foreign currency losses in the third quarter, was around TWD 3.6 billion.

So the net income attributable to the stockholder of the parent in 2018 was TWD 7 billion or an EPS of TWD 0.58, a decline of 26% year-over-year.

Cash has continued to increase and now at the end of 2018, was TWD 83.6 billion.

Free cash flow alone in 2018 was a bit over $1 billion U.S. And stockholder equity in 2018 was TWD 206 billion.

As I mentioned, in Q4 blended ASP declined in a more notable range, mainly because of our 14-nanometer revenue has been suffering very weak demand from cryptocurrency related customers.

In Q4, our geographic breakdown wasn't changed much.

So North America went up to 38% from 34% in the previous quarter.

And the rest of the regions declined accordingly.

And for the full year, Asia still remained our biggest share of revenue breakdown, around 50%.

And North America declined year-over-year to 38%.

IDM stayed around 7% to 8% in both quarters.

And the whole year didn't change much, at still around 8% to 9%.

And communications is 44% of the revenue breakdown.

Consumer is 30% in Q4.

And for the full year, communications declined by 4 percentage points to 45% and computer and others increased marginally.

Our 14-nanometer revenue, as I mentioned earlier, has come down significantly to 1% of the revenue, from 5% in Q3, mainly due to the volatility in the cryptocurrency market.

28-nanometer also declined to 10% from 13% in the previous quarter.

And the rest of the technology nodes remained relatively similar.

And for the full year, 14-nanometer increased from 1% to 3% and 28-nanometer decreased from 15% to 13%.

For the quarterly capacity, we have factored in the annual maintenance in the Q1 forecast.

So most of the fabs, except for HeJian, will show a decline in available capacity, mainly due to the annual maintenance.

Our expected or budget CapEx for 2019 is about $1 billion U.S. Our actual CapEx for 2018 was around $650 million U.S.

And that's the summary of UMC's results for Q4 2018.

More details are available in the report, which has been posted on our website.

I'm now turning the call over to our President, Mr. Jason Wang.

Thank you, Chitung.

Good evening, everyone.

Here I'd like to update the fourth quarter operating result of UMC.

In 2018, we started seeing the early fruit of our strategy with measurable results.

Our disciplined capital expenditure approach helped to generate a free cash flow total of TWD 31.34 billion for the year.

In addition, we completed 2 rounds of a treasury share buyback for cancellations, amounting to approximately TWD 6.5 billion.

Looking into the first quarter of 2019, we anticipate further deceleration in customers' wafer demand due to a softer than expected outlook in entry-level and mid-end smartphones, as well as falling cryptocurrency valuation.

Although UMC's ongoing transformation will need time to reach its full synergy and potential, our progress so far has enabled the company to better endure these current headwinds.

Going forward, we'll continue executing our strategy of evaluating and pursuing return-driven investment, while focusing on our technology strengths within specialty processes on existing nodes.

We are confident that our sustained effort and calculated global capacity expansion will strengthen UMC's resilience during the challenging market, while favorably position the company to take maximum advantage during a strong demand cycle.

Before I get into Q1 guidance, I would also like to update you on the ongoing legal cases involving UMC.

To clarify UMC's position, UMC has and will continue to comply with all applicable laws and regulations, and we have suspended all R&D activities we are performing for Fujian Jinhua.

Since being established 39 years ago, UMC has been an indispensable player in the global supply chain, with volume production technologies down to an advanced 14 nanometers.

From 1996 to 2010, UMC accumulated nearly 15 years of experience in manufacturing DRAM products.

At one point, UMC's internal DRAM team had well over 150 people.

Thanks to its extraordinarily stable workforce, UMC has ever since possessed and preserved a wealth of DRAM knowledge and experience as an institution.

The joint development project under which UMC agreed to develop the DRAM process for Jinhua, which it was a standalone project entirely separated from UMC's pure-play foundry services, was nothing but a pure business transaction that made all the business sense for UMC at the time.

It was duly submitted to the Taiwan authorities, which approved the project in its entirety in April 2016.

Notably, that was a time when the U.S.-China trade war was unheard of.

UMC wants to ensure our customers and stakeholders that UMC will vigorously defend itself against all false charges and misconceived allegations.

Now let me move on to the first quarter 2019 guidance.

Our wafer shipments will show a decrease of 6% to 7%.

ASP in U.S. dollars is expected to decline by 1% to 2%.

Gross profit margin will be in the mid-single percentage range, and capacity utilization rate will be in the low 80% range.

For foundry CapEx budget of 2019, it will be $1 billion U.S.

That concludes my comments.

Thank you all for your attention.

Now we are ready for questions.

(Operator Instructions) Your first question is coming from Randy Abrams, Credit Suisse.

Yes, I wanted to ask on the first question for the CapEx where you spent below the budget at $650 million for 2018.

Could you talk for 2019 the budgets coming back up to $1 billion despite lower utilization to start the year?

So could you go through a bit more on the driver for the increase and where you plan to spend or allocate that $1 billion?

The allocation for the $1 billion budget is about 75% in 12-inch and 25% in 8-inch.

But as you are aware, this is actually a small number in absolute terms.

And last year, our budget was around $1 billion as well.

But it ended up some of the payment and some of the equipment delivery, et cetera.

So it's a small absolute dollar being deferred into 2019 budget.

So in a sense, it's a trend down.

It's a downward trend.

But in terms of the actual payment terms, there may be some kind of small magnitude of fluctuations.

Okay, and maybe a follow-up to that question, if you could talk to the capacity now, I think the HeJian plant to add with that 25% to 8-inch; and then also the 75% for 12-inch, is there any capacity for specific nodes or is it more technology upgrades within it that may not necessarily be net wafer capacity additions?

Yes, the site at 8-inch in HeJian, the 12-inch is mainly [spending in 12 IR Singapore] facility and the technology node we're increasing is 65 nanometers node.

Okay, and can you say like, how much capacity in Singapore, or is it just upgrading the line to more advanced or different process?

It's mostly upgrade and there's some automation involved in both Taiwan and Singapore as well.

So again, this is small absolute numbers.

So it's not much capacity increase, let's say, in 2019.

Okay, and I wanted to ask a question just on the gross margin and the cost structure.

I guess for the gross margin, the decline from mid-single digit, maybe aside from the shipment, I think sometimes at the beginning of the year there's a step down in pricing.

But your ASPs are only down slightly.

So I guess first, maybe talk about the driver to bring it to mid-single digit, and then if there's any mix change you're expecting, say, between nodes, like if advanced capacity is maybe holding up or how you see the relative mix between nodes.

The drop in the Q1 gross margin, the leading contributor to the lighter utilization as you said, is if we project it from 88% down to a low 80% range.

And if you break it down by the technology node, the 28 stays pretty flat, and mainly drop is coming off from the 40-nanometers.

But in addition to the lighter utilization, the annual tool and equipment maintenance which we normally execute in Q1, along with fewer working days, that's also a reason that contributed to that decline in Q1 gross margin right now.

Okay, and final question, if I could just elaborate on the 40-nanometer, in the remarks I guess you had talked some about smartphone, low and mid-end smartphone.

So maybe within 40, is it more of a kind of short-term market weakness or more broadly how you see sustainability or outlook for the 40-node, whether you're seeing customers migrate to 28, and if you expect to retain or capture that if you are seeing that, like where you'll start to see the 28 ramp-up?

Yes, I mean for the 40, we do see the technology continue to migrate.

We expect 40-nanometer loading will improve, due to the high voltage ultra-low power logic and the non-volatile memory will come in, and as we continue to see the 40-nanometer demand shift over to this area, from AMOLED display to IoT devices.

And as a result if we look at UMC alone, our 40-nanometer penetration rate in AMOLED display and IoT area are increasing.

Moreover, we have been working on the migration plan for the current 55-65 logic application, moving towards to a 40.

So we have some applications coming over from 55-65 over to 40.

Therefore, we believe that 40-nanometer loading weakness will be a temporary issue.

We are seeing more tape-out on the 40 in 2019 from last year.

There are some sizeable increases in year-over-year rate.

So we're fairly confident that 40 nanometers loading will recover gradually.

Okay, and non-volatile memory, is that embedded flash or is that you'll start to do -- you have any standalone flash memory in that?

No, it's embedded flash, not multi, the embedded.

And the next one is from Gokul Hariharan from JP Morgan.

First of all, could you comment a bit on the shape of the inventory correction that you're seeing?

I think you've probably had now 2 quarters of revenue decline.

Where do you expect the inventories to bottom out?

Should we expect that Q2 is also going to be a quarter where we kind of need to see some inventory correction?

I think your larger competitor had mentioned that they're expecting inventories to bottom out only by the middle of the year.

That's my first question.

And could you also give us a little bit more color on the 8-inch supply-demand situation, given we were pretty tight in the first half of last year?

What has been happening on the 8-inch side from a supply-demand perspective?

Sure.

I'll start off with the 2019 outlook.

From a year, the entire year-wise, the market research is showing the foundry [goals] will be flat this year.

And starting from this Q1, actually starting from the Q4; we already start observed market weakness and inventory correction within the entire semiconductor supply chain.

And on top of that, the U.S.-China trade dispute has created additional uncertainty on a micro level.

So we believe that 2019 will be a challenging year.

So as far as going into Q2, I think at this point it's a bit early to comment.

So for your second question about the 8-inch, the 8-inch market also is experiencing smartphone demand weakness and inventory correction in the supply chain as a short term.

Our 8-inch business was still running close to 100% utilization in Q1.

So we do see some impact, but it's marginal.

It's small.

So we remain positive for the long-term 8-inch business, as we're seeing more diverse 8-inch demand from power IC, MCU, automotive, IoT devices.

With our more comparative specialty technology in those areas, we are confident that we can address those 8-inch demands well.

Okay, so just to belabor the point a little bit, so as you mentioned, we have been in inventory correction.

Do you feel that exiting Q1 your customers would have largely worked down the inventory or do you still think that there is going to be some excess inventory for your customers exiting Q1?

Well, we see a sudden stop in inventory -- in building additional wafers.

So we start seeing an inventory build-up in the entire supply chain.

I think to expect that Q1 would digest those build-up will be optimistic.

And at this point, given the macroeconomic uncertainty, I think it will last more than Q1.

But again, beyond Q1 it's too early to comment at this point.

Okay, great.

Can I ask a quick question?

Any change for the China subsidy that you get after the changes in the DRAM development program?

Is there any impact on the China subsidies that we were recognizing and we expect to recognize over the next couple of years?

These are totally 2 different businesses.

The DRAM is pure technology development.

We're receiving in NRE fee in return to develop DRAM for Fujian Jinhua.

Our subsidiaries, including USE in Xiamen and also the wholly owned subsidiary of [DNH] is part of UMC's foundry service and we serve diverse foundry customers and try to focus on Chinese upcoming foundry customers, but also serve customers outside of China as well.

So those are 2 totally different business models.

And the current issues with Jinhua has nothing to do with UMC's foundry business.

And then next one is coming from Szeho Ng from China Renaissance.

My first question is regarding the Fujitsu fab.

When should we expect the fab to be fully consolidated into the group?

The Fujitsu Mie fab, the acquisition, is still pending.

So we both, between Fujitsu and UMC, are working toward to closing this Mie fab.

So we are not consolidated yet.

I see.

But when should we expect the closure to happen?

Because previously the companies cited closure to be the beginning of this year, in January.

Right.

That was the original target.

It was January 1. And the current status is that our equity transaction was approved by the ROC government September 26 last year.

However, the transaction has been partially delayed pending approval from certain relevant government authorities.

So as a result, that will push out to the next available date.

And right now we're targeted for April 1.

Okay, got you.

Okay, thank you.

And second question, when I look at the company's P&L, for the last couple of quarters, the company actually gets quite a big net other operating income.

Should we expect a similar thing to happen going into 2019 at least?

Yes, for 2019, which is mainly coming from subsidies of our investment in both China and other areas.

Those are likely to be in the same range for most of 2019.

And the next question is coming from Sebastian Hou from CLSA.

My first question is, can you comment on your utilization rate for 8-inch and 12-inch in fourth quarter '18 and 1Q '19?

The Q4 '18, the 8-inch is running over 100%.

And the 12-inch is about mid-70s, and overall it's about 88%.

For the Q1 '19, we're saying the 8-inch will be close to 100% and the 12-inch will be in the low 70s.

So overall will be somewhere in the 80% range.

Okay, thank you.

My second question is on the -- I think Chitung already mentioned that you recognized a one-off OpEx related to the DRAM R&D.

So can you get the subsidy NRE from Fujian Jinhua later on?

The onetime increase in our OpEx of TWD 700 million which related to the DRAM development project, about TWD 406 million is lost credit, which is we're supposed to collect the money, as we already reached the milestone.

But the receivable appears to be in risk for the current status.

So that's why we recognized TWD 406 million lost credit.

So that's the one you referred to.

And in the future, there's no other things like this going forward.

And the remaining TWD 300 million plus is mainly due to the cost we already put into the project.

Again, with the current status of the DRAM project, it's unlikely to recover.

So we also recognized that as an expense at one time in Q4 of 2018.

And that also is unlikely to repeat again in Q1.

Okay, got it.

Thank you.

Do you sense any change to your relationship with your U.S.-based customers?

As I'm not sure if you sense any of your U.S.-based customers may be concerned regarding your lawsuit you're involved with U.S. government right now.

Well, for that legal case, well first of all, UMC does not manufacture or sell any DRAM products.

So this recently filed U.S. case does not impact either our customers' production or our delivery schedule.

So we haven't seen any impacts on that front yet.

And so UMC business continues as usual for the pure foundry side.

Okay, and my first question is I recall there is a Fujian Jinhua and your DRAM projects with Fujian Jinhua has also involved some R&D equipment that plays in UMC factory in Tainan.

So how will those equipment be -- how are we going to deal with those equipment and where would those equipment be disposed or how to deal with them later on?

Well, right now the project has been suspended, as we reported, and the tools are not being used.

In other words, the whole equipment dedicated for the DRAM has been suspended at this point.

And we're just not touching it.

Okay, so the tools just sit there idled and you don't have to be responsible for any depreciation of that either?

No, not at all.

No.

Okay.

Can you resell it or you just leave it there?

That's not ours.

It's not our equipment.

Okay, all right.

So they will be taken back at some point later?

We don't know.

They're starting the process, the judicial process.

So it's too early to comment about the future of the equipment.

Right, right.

The last question from me is on the dividend policy that given that you have generated a lot of cash flow in '18, and what's the dividend policy -- what would the dividend policy be like for this year?

We will continue with a very high payout ratio, and again, UMC's really coming from a total shareholder return approach.

Other than a very high payout ratio in cash dividend, we also conduct 2 rounds of share buybacks and for cancellations.

So it's going to be a combination effort for UMC to go for this maximum shareholder return approach.

Okay, so Chitung, when you mention high payout ratio that means it's likely to be over 100%?

It's not impossible, but again, we want predictable and also steady flow, and hopefully a long ways out enhanced increased earnings.

So it's not again only focused on the dividend, cash dividend.

We want to come from several other means to enhance our total shareholder return.

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

So my first question is to follow up the U.S. case, right?

So does the company need to prepare any provision loss for the potential results and whether it would impact the coming quarters' OpEx?

Well, the case is at a preliminary stage, where all amounts at this point are speculation.

For the final damages, it's premature.

So given the current process, we are not booking anything yet.

Okay.

And for full year, I know visibility is quite low, even to second quarter.

But does the company set any full year target in terms of revenue and also the gross margin?

We are not allowed to give full year revenue or gross margin guidance, according to the local regulators.

However, UMC certainly will try to defend and enhance our position in the foundry sector.

We hope we will be able to grow in line with the foundry sector for the 2019.

Okay, that's fair.

So I guess 28-nanometer is kind of a big wild card.

It has been slow for the past 3 quarters.

Does the company expect that 28-nanometer utilization can improve in the coming quarters?

Well, right now in Q4, we continue seeing the weakness of smartphone and automotive demand we kind of touched earlier.

So we do not anticipate it will improve in Q1.

Okay, but at the same time there is a kind of overcapacity issue for 28-nanometer.

So how do you think about the coming competition in the coming quarters?

Even the demand comes back, does the company have the confidence to maintain the current market share?

Well, I mean our current market share is not -- I mean our goal is to grow, not to maintain our current market share because market share is relatively low.

But if you would look at the 28 nanometers, we observe the evolution of the 28 and 22 high-k derivative technology are reaching to an industry standard now.

So in the past, we've mentioned we've been seeing quite a bit of derivative for the past couple years.

So as [states] are reaching to the industry standard, which will significantly help us to capture better business opportunity in the market shares.

So in long term, we're actually seeing positive growth in this area, both from demand and our readiness point of view.

Okay, and maybe also a question about the cost structure, right?

So the part one is about your depreciation trend in 2019, and also I think recently you guys have started talking about foundries should cut costs at those low wafers, et cetera.

So can the company give us some color about the raw wafer price chain in 2019?

To answer your first question, UMC's depreciation trend is down about 5% plus in 2019 versus 2018.

2018 itself declined about 2% compared to the previous year.

For your second question, in terms of the raw wafer pricing situation, last year in 2018, the importance of the raw wafer is to secure long-term contract agreements with the suppliers.

So although we have secured that long-term contract agreement with the raw wafer suppliers, we will continue to negotiate with the supplier to achieve a better cost savings.

So this is an ongoing effort, yes.

Yes, so again, can you clarify is this kind of a negotiation is for 2019 contracts or for those contracts in the following years?

It depends, but I think the overall negotiation will continue.

And whether it's going to apply to the current contract or the next-year contract, we're still in negotiation with the supplier.

Okay, thanks.

And sorry for those tough questions.

And actually I see one kind of bright spot is about your 8-inch business, right?

So can management help us to understand?

Because essentially lots of 8-inch end markets like automotive, industrial, even starting to show weakness, right?

And those IDM customers tend to do an in-source during a downturn.

So can you give us some color why UMC can still maintain such high utilization for 8-inch and can you give us some kind of a long-term demand-supply perspective for the 8-inch business?

Thanks.

Well, for the 8-inch, long term we are optimistic about the outlook in the long term.

There's still going to be many applications drives the 8-inch needs demand, power IC, MCU, automotive, IoT devices.

And so I'm saying there's still going to be a very diverse application drives the 8-inch demands.

And overall, 8-inch is still under constraint.

So we think it will remain healthy for the 8-inch.

And so the key focus here is to provide competitive specialty technologies.

So that will be our focus.

We will be focused on delivering those specialty technologies in our RF SOI, BCD, embedded, non-volatile, and as long as we can deliver those competitive solutions then we have confidence to address those 8-inch demands in the long term.

Okay, how about the big segment, large panel drive ICs; that kind of a big kind of end market for 8-inch?

How does it look for 1Q for the wafer demand?

Well, the overall market is showing some weakness.

But I don't like to comment specific each application, but for the Q1 we see across the segment weakness.

And lastly, just one very last question; so Chitung, can you give us an OpEx range for 1Q?

Because I know there were some onetime items, right?

But what was 1Q OpEx range?

I think it should be the Q4 number, but it's currently a onetime charge.

And the next one is coming from Julie Tsai from UBS.

I have two questions.

One is I just want to clarify Q1 guidance of gross margin.

Did you say mid-single digits?

This is compared to Q4's 13%, correct?

But the utilization rate doesn't seem to drop so much, and wafer and ASP decline in Q1 also doesn't seem to be that bad.

Could you give us a little bit more guidance on that kind of gross margin indication?

Well first, yes.

The answer is yes.

We're dropping from 13% to mid-single digit.

And the low utilization from 88% drop down to low 80s, you can probably roughly looking, it's about 4% impact, 4 point impact to the gross margin, when we see a utilization drop of 8% or in the range of that.

As I reported earlier, our annual tool and equipment maintenance which we normally execute in Q1; that will probably give you another point, and as well is the fewer working days.

So a combination of those will probably impact us that much.

I see.

Okay, and also second question is that it seems like UMC is still aiming to develop more advanced technology, for example, 14 nanometer.

Or are you still keen to stay at more mature nodes where you are quite competitive compared to other foundry players?

Could you give us a bit of guidance on that?

Well, our strategy is a focus on the mature technology, specialty technology.

And our strategy is to try to be focused and be relevant in those areas.

So we see many different applications drive the technology needs in the specialty area, so not particularly in the 14.

Our current R&D resources are mainly focused on the diversity of the technology area.

So we'll continue developing our solution across the multiple technology nodes to address, for example, power IC or as mentioned earlier the display driver, RF switch, MCU application.

And we believe to provide competitive specialty technology solution to our customer will improve our product mix and sustain our overall utilization rate in line with our capacity expansion plan.

We don't really highlight the advance of 14 in the past few quarters already.

But we already developed the 14, and we will finish the 14.

But we will probably stop at 14 and we have no plans to go beyond 14 at this point.

I think that's much clearer.

And can we also assume that maybe a CapEx of $1 billion, that's probably fairly high or good enough for maybe this year.

And next year and beyond would be lower than this?

No, we don't give CapEx guidance beyond 2019.

2019 it's about $1 billion.

Last year CapEx, some of them, part of them fall into this year.

And so we believe this is currently a good number for us and we continue to accumulate cash on the back of this disciplined CapEx approach.

(Operator Instructions) And next we'll have Randy Abrams from Credit Suisse.

Yes, I just wanted to see with the Fujitsu now planned April 1, if you could give maybe a rough range for sales and profitability or an OpEx run rate, just to factor in once you do get the closure?

It's a part of Fujitsu [full year], we are not allowed to comment on the outlook unfortunately.

Okay, I guess we'll have to wait until next quarter.

A follow-up then on the OpEx where I think exit charges, it's running flat.

I think medium term with some of the move now to do more of like a CapEx-light and more approach on returns, could you maybe discuss kind of your view on OpEx, if stable is the way to think about it or there's still areas you need to invest and grow, or kind of direction we should think about OpEx kind of over the next 1 to 2 years?

In the next 1 or 2 years, our goal is to have a stable percentage ratio of revenue.

Longer term, of course, is the management's job to further -- to utilize those resources more efficiently.

So hopefully longer term, we will have it under control.

But in the near term, in 2019, given the challenging outlook in terms of the whole market, our goal is to maintain a stable percentage of revenue for the OpEx ratio.

Okay, and the final question on the 28, it came down a bit more to 10% of revenue.

And I think in the past couple quarters you mentioned that's now kind of a larger range of tape-outs, but smaller volume.

When you look at those in the production, if you could give a sense how we should see the 28, if those tape-outs start contributing to grow that as a percentage of sales to improve loadings or remain kind of in this range for a little longer?

Yes.

A number of the 28 tape-outs continues to grow in both communications and computer [standards].

And if I break down the ratio, we are seeing a 70% on tape-out becomes a high-k solution.

You can see that.

That's a big improvement.

However, just like you said, the 28-nanometer high-k still composes a fragment of order customers from the second and third wave application.

So the ramp-up time will still take much longer.

But I think we definitely see the tape-out momentum now.

Okay, but I guess that meaning by taking longer, probably stable?

Like, we're at a level now, it probably stabilizes for a while and then probably later in the year can start to grow -- can start to get back to growth again?

That will be our expectation, yes.

It appears to be no further questions at this point.

So that will be the end of our Q&A session.

I'll turn it over to UMC's Head of IR for closing remarks.

Go ahead, please.

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.

And ladies and gentlemen, that concludes our conference for fourth quarter 2018.

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, Events section.

You may now disconnect.

Good-bye.