台積電 ADR (TSM) 2020 Q2 法說會逐字稿

(foreign language) Ladies and gentlemen, welcome to TSMC's Second Quarter 2020 Earnings Conference Call.

This is Jeff Su, TSMC's Director of Investor Relations and your host for today.

To prevent the spread of COVID-19, TSMC is hosting our earnings conference call via live audio webcast through the company's website at www.tsmc.com, where you can also download the earnings release materials.

(Operator Instructions)

The format for today's event will be as follows.

First, TSMC's Vice President and CFO, Mr. Wendell Huang, will summarize our operations in the second quarter 2020, followed by our guidance for the third quarter 2020.

Afterwards, Mr. Huang; and TSMC's CEO, Dr. C.C. Wei, will jointly provide the company's key messages.

Then, TSMC's Chairman, Dr. Mark Liu, will host a Q&A session, where all 3 executives will entertain your questions.

As usual, I would like to remind everybody that today's discussions may contain forward-looking statements that are subject to significant risks and uncertainties which could cause actual results to differ materially from those contained in the forward-looking statements.

Please refer to the safe harbor notice that appears in our press release.

And now I would like to turn the call over to TSMC's CFO, Mr. Wendell Huang, for the summary of operations and current quarter guidance.

Thank you, Jeff.

Good afternoon, everyone.

Second quarter revenue was flat sequentially, as the continued 5G infrastructure deployment and HPC-related product launches offset weaknesses in other platforms.

Gross margin increased 1.2 percentage points sequentially to 53% mainly due to continuing high level of utilization and the absence of unfavorable inventory valuation adjustment, partially offset by NT dollar appreciation in the second quarter.

Total operating expenses increased by TWD 1.19 billion, mainly as TSMC supported a range of COVID-19 relief efforts.

Operating margin increased by 0.8 percentage points sequentially to 42.2%.

Overall, our second quarter EPS was TWD 4.66, and ROE was 28.5%.

Now let's move on to the revenue by technology.

7-nano process technology contributed 36% of wafer revenue in the second quarter, while 16-nanometer contributed 18%.

Advanced technologies, which are defined as 16-nanometer and below, accounted for 54% of wafer revenue.

Moving on to revenue contribution by platform.

Smartphones decreased 4% quarter-over-quarter, to account for 47% of our second quarter revenue.

HPC increased 12%, to account for 33%.

IoT decreased 5%, to account for 8%.

Automotive decreased 13%, to account for 4%.

Digital consumer electronics decreased 9%, to account for 5%.

Moving on to the balance sheet.

We ended the second quarter with cash and marketable securities of TWD 605 billion.

On the liability side, current liabilities increased by TWD 25 billion mainly due to the increase of TWD 30 billion in short-term loans.

On financial ratios, accounts receivables turnover days increased 2 days to 44 days.

Days of inventory also increased 2 days to 55 days mainly due to N5 ramp and stronger N7 demand.

Now let me make a few comments on cash flow and CapEx.

During the second quarter, we generated about TWD 170 billion in cash from operations, spent TWD 127 billion in CapEx and distributed TWD 65 billion for third quarter cash dividend.

We also increased TWD 30 billion in short-term loans and issued TWD 36 billion of corporate bonds.

Overall, our cash balance increased TWD 37 billion to TWD 468 billion at the end of the quarter.

In U.S. dollar terms, our second quarter capital expenditures amounted to $4.2 billion.

I've finished my financial summary.

Now let's turn to our third quarter guidance.

Based on the current business outlook, we expect our third quarter revenue to be between USD 11.2 billion and USD 11.5 billion, which represents a 9.3% sequential increase at the midpoint.

Based on the exchange rate assumption of USD 1 to TWD 29.5, gross margin is expected to be between 50% and 52%, operating margin between 39% and 41%.

Now I will hand over the call to C.C. for his key messages.

Thank you, Wendell.

Good afternoon, ladies and gentlemen.

Let me start with our near-term demand outlook.

We concluded our second quarter with revenue of TWD 310.7 billion or USD 10.4 billion, in line with our guidance given 3 months ago.

Our second quarter business increased slightly in U.S. dollar terms, as the continued 5G infrastructure deployment and HPC-related product launches offset weakness in other platforms.

Moving into third quarter 2020, we expect our business to be supported by strong demand for our industry-leading 5-nanometer and 7-nanometer technologies driven by 5G smartphone, HPC and IoT-related applications.

Looking at the second half of this year.

COVID-19 continues to bring some level of disruption to the global economies, and uncertainty remain.

We have observed weak consumer demand in the first half of this year and now expect global smartphone units to decline low-teens percentage year-over-year in 2020.

However, amid the COVID-19 pandemic, we also observe the supply chain making effort to ensure supply chain security and actively preparing for new 5G smartphone launches.

We raised our forecast for 5G smartphone penetration rate to high-teens percentage of the total smartphone market in 2020.

For the full year of 2020, 5G and HPC-related applications will continue to drive semiconductor content enrichment.

We now forecast the overall semiconductor market, excluding memory growth, to be flat to slightly increasing, while foundry industry growth is expected to increase to be mid- to high-teens percentage.

For TSMC, although COVID-19-related uncertainties remain, our technology leadership position enable us to outperform the foundry revenue growth.

We believe we can grow above 20% in 2020 in U.S. dollar terms, including the impact from the new U.S. regulations, which I will discuss in the next session.

Our 2020 business will be supported by strong demand for our industry-leading 5-nanometer and 7-nanometer technologies and our specialty technology solutions, driven by customers of 5G smartphone-related product launches and expanding HPC-related opportunities.

Now let me talk about the impact of new U.S. regulations.

On May 13, the U.S. Department of Commerce announced a set of new export control regulations.

As a global and law-abiding company, TSMC will follow all the rules and regulations fully, no doubt about it.

While there may be some impact from the new U.S. regulations, TSMC's purpose to unleash innovation remain unchanged.

Our leading position in the semiconductor industry [as everyone's foundry], technology leadership, manufacturing excellence and customers' trust also remain unchanged.

We will continue to build upon our trinity of strengths and conduct our business with integrity to ensure our value and contribute to the semiconductor industry.

In the near term, we will work dynamically with our customer to minimize the impact to our business from new U.S. regulations.

In the mid- to long term, we believe the underlying megatrend of 5G-related and HPC applications remain intact, and supply chain can adjust and rebalance themselves.

With our technology leadership, we are well positioned to capture the mid- to long-term growth opportunities.

We reaffirm our goal to grow at the high end of our long-term growth projection of 5% to 10% CAGR in U.S. dollar terms.

Next, let me talk about our N5 ramp-up and N4 introduction.

N5 is the foundry industry's most advanced solution with the best PPA.

N5 is already in volume production with good yield, while we continue to improve the productivity and performance of the EUV tools.

We are seeing robust demand for N5 and expect a strong ramp of N5 in the second half of this year driven by both 5G smartphones and HPC applications.

As we observed some delays earlier this year in N5 toward deliveries due to COVID-19, we now expect 5-nanometer to contribute about 8% of our wafer revenue in 2020.

We also introduced N4 as an extension of our 5-nanometer family.

N4 will have compatible design rules and a highly competitive performance to cost advantages, as compared to N5, and we'll target next wave of the N5 products.

Volume production is targeted for 2022.

Thus, we are confident that our 5-nanometer family will be another large and long-lasting node for TSMC.

Now I will talk about our N3 status.

N3 will be another full node stride from our N5 with about a 70% larger density gain, 10% to 15% speed gain and 25 to 30 power improvement, as compared with 5-nanometer.

Our N3 technology will use FinFET transistor structure to deliver the best technology maturity, performance and cost.

Our N3 technology development is on track, with good progress.

N3 risk production is scheduled in 2021, and volume production is targeted in second half of 2022.

We have already demonstrated 256 megabit SRAM functionality.

N3 logic test chip is fully functional, with yield ahead of plan.

The device performance is also on track.

Our 3-nanometer technology will be the most advanced foundry technology in both PPA and transistor technology when it is introduced, which will further extend our leadership position well into the future.

Finally, let me talk about our U.S. fab plan.

On May 15, we announced our intention to build an advanced semiconductor fab in the U.S. We have received a commitment to support this project from both the U.S. federal government and the State of Arizona.

We are working closely with them as well as our supply chain partners to build an effective supply chain and make up the cost gap.

This fab will start with 5-nanometer technology with 20,000 wafer per month capacity.

Production is targeted to begin in 2024.

The U.S. fab will enable TSMC to expand our technology ecosystem and better service our customer and partners.

At the same time, as TSMC global presence increases, it will allow us to better reach global talent to sustain our technology leadership.

Now let me turn the microphone over to our CFO.

Thank you, C.C.

Let me start by making some comments on our second half profitability outlook.

We have just guided third quarter 2020 gross margin to decline by 2 percentage points sequentially to 51% at the midpoint primarily due to the margin dilution from the initial ramp-up of our 5-nanometer technology in the third quarter and a less-favorable foreign exchange rate.

As compared with our expectation 3 months ago, our third quarter gross margin midpoint is higher, mainly supported by a high level of overall capacity utilization despite the uncertainty from COVID-19.

Looking ahead to the fourth quarter, we expect the continuous steep ramp-up of our 5-nanometer to dilute our fourth quarter gross margin by about 2 to 3 percentage points.

Now let me talk about our capital budget for this year.

Every year, our CapEx is spent in anticipation of the growth that will follow in the next few years.

While the impact of COVID-19 virus brings uncertainties in 2020, we have seen our business holding up well, so far, thanks to our technology leadership at 5- and 7-nanometer nodes.

Looking ahead, the multiyear megatrends of 5G-related and HPC applications are expected to continue to drive strong demand for our advanced technologies in the next several years.

In order to meet this demand and support our customers' capacity needs, we have decided to raise our full year 2020 CapEx to be between USD 16 billion to USD 17 billion.

We also reiterate that TSMC is committed to sustainable cash dividends on both an annual and quarterly basis.

That concludes my key messages.

Thank you, Wendell.

This concludes our prepared statements.

(Operator Instructions) Should you wish to raise your question in Chinese, I will translate it to English before our management answers your question.

(Operator Instructions) Now let's begin the Q&A session.

Operator, please proceed with the first caller on the line.

The first to ask questions, Gokul Hariharan, JPMorgan.

Great results in a tough time.

Just a quick question on how we think about N3's development.

Do we feel that N3 -- since we're talking about mass production in the second half of 2022, usually, the new nodes start sometime in Q2.

I just wanted to understand.

Are we thinking about a slightly slower ramp for N3 compared to what we've had in the first year for N5 as well as N7?

That is my first question.

My second question is when we think about leading edge, once the U.S. regulation starts to come in, how do we think about managing capacity?

Do we feel that the capacity can get filled up relatively quickly once, one of our leading customers, you have to start shipment to them?

Or do we feel that there could be a couple of -- there could be some time where there could be a little bit of underutilization?

Okay.

Thank you, Gokul.

Let me try to -- allow me to summarize your question.

Your first question is related to N3.

How do we think about the N3 development?

We have said the mass production timing is in second half '22 versus typically the second quarter, so should we expect a slightly lower ramp of N3?

This is your first question.

Okay.

Let me answer that, Gokul.

In fact, we develop our new leading-edge technology.

We work closely with our customer.

So the schedule and also the ramp-up, also the progress, we're all working with customer closely and determine when and -- to be the best timing.

So far, our N3 development is very smooth and successful.

And we still target the risk production in next year and ramp-up in the second half.

There's -- all the schedule is working with our customers.

Okay.

And then, Gokul, your second question is on the leading edge and in light of the recent U.S. regulations.

How will we manage our capacity at the leading edge?

Will we see a gap in the utilization?

Or will we be able to fill it up?

It should be no problem because of -- as we just stated, that the 5Gs megatrend and also HPC-related application continue to be very strong.

And we observed that all our customer are very actively prepared for these 2 application, 5G and HPC.

In addition to that, we also observed that our customers are -- try to secure their supply chain security and which is very important with this COVID-19's uncertainty.

Let me add...

Do we, do you feel that even for N5 that is applicable?

Or...

Yes.

Even with N5, yes.

Let me add to that.

I think, for the short term, some impact is inevitable.

Currently we work closely with a customer very dynamically, trying to fill up the capacity.

And for the long term, as C.C. mentioned, we are very -- we're still optimistic.

Okay.

Thank you, Gokul.

Next to ask questions, Sebastian Hou, CLSA.

So my first one is I wanted to get some -- pick your brain about how do you evaluate the feasibility and possibility of building up an advanced node fab without American contents, be it technology equivalent IP material, et cetera, in the next 5 years or 10 years or even longer.

Is it worth it?

Or even if it takes a long time and tremendous efforts, would TSMC ever consider that?

Well, let me answer that question.

We know that a U.S. fab, as compared with a fab in Taiwan, the cost structure is actually a little bit higher.

And that's why we say that we are working with the federal government and also the State of Arizona to close the gap.

Yes.

Sorry.

Just to repeat the question.

I think Sebastian's question is asking about building up an advanced node fab or production line without using any so-called American contents whether in terms of equipment, technology or IP materials.

He wants to know, in the next 5 to 10 years, is it feasible?

Is it worth it?

And is this something that TSMC would consider?

Let me pick up this one here.

The semiconductor technology is very unique in this industry.

The technology continue to improve.

Every 2 years, there will be a new generation of technology come out to serve the best-performance product.

And therefore, we -- I think our main force is still pursuing the technology leadership, trying to overcome each generation's challenge.

And to do that, I think our current focus is still working with our equipment partners, dealing with -- utilize the best of the kind equipment that we can have to pursue our business growth.

So you're right, to -- if we do that otherwise, technology advancement will be extremely challenging.

It will be extremely difficult, not to talk about 5 to 10 years alone.

So that is not our current effort at this point.

Okay.

Sebastian, do you have a second question?

Yes, yes.

Second question is I'd like to follow on the -- ask about the inventory situation.

The first, could you -- can you update us on how you see the fabless days of inventory at the end of Q2 and how you see that in the second half this year?

And also on the inventory side, it looks like it's getting increasingly difficult to look at the inventory from a comprehensive perspective.

Fabless DOI may not be enough because apparently there is a lot of the Chinese companies stockpile the inventory in fear of being sanctioned.

And also, across the board globally, the whole supply chain has been raising the [fixed store level] inventory in the past few months in fear of supply chain disruption caused by COVID-19, so -- but those are not reflected in fabless DOI.

So how do we see about this inventory and potentially hidden excessive inventory situation going forward, which could -- do you concern about that to be a potential overhang at some point the destocking could come?

Okay.

Let me summarize your second question, Sebastian.

Both of it relates to the inventory situation.

The first part is what is -- for TSMC tracking of fabless customers, what is the fabless DOI exiting 2Q and the outlook into second half?

That's the first part of your question.

And then your second part of your question is are we concerned that the inventory may -- situation may see some hidden or discrepancies due to whether it's COVID-related supply chain disruption or the U.S. regulation and such?

Will this lead to a hidden inventory risk?

And is there a risk of inventory correction?

Okay.

Let me answer that.

The inventory level of our fabless customers that we track exited first quarter above the seasonal level.

We expect a further increase in second quarter and then stay at a high level in the second half, as the supply chain is making efforts to ensure supply chain security and our customers are in high anticipation and preparing for new 5G smartphone product launches in the second half of this year.

We cannot rule out the possibility of an inventory correction sometime down the road.

We observe the supply chain active making efforts to ensure the securities and active preparation for 5G smartphone launches.

We will just have to wait and see how the sell-through goes.

Okay.

Thank you, Sebastian.

The next caller is Bill Lu from UBS.

Yes.

I'm wondering if you can comment on the CapEx guidance for this year.

It's now raised to USD 16 billion to USD 17 billion.

I'm wondering what that increase is, whether it's 5-nanometer or it was something different.

Secondly, related to that, can you talk about your CapEx intensity structurally, whether this increase is temporary and whether this is pulled in from next year, therefore maintaining the longer-driven intensity?

Or how we should think about that.

Okay.

Let me summarize your 2 questions, Bill.

Your first question is in relation to our 2020 CapEx guidance and the range of USD 16 billion to USD 17 billion.

So Bill wants to know what is driving this increase and then secondly, in terms of the capital intensity outlook over the next few years.

Okay.

The CapEx increase from 3 months ago for this year is basically comes from the advanced technologies.

And the capital intensity this year will be slightly lower than 40%, and over the long term, it will gradually go down to about mid-30s.

Okay.

Thank you, Bill.

The next caller is Brett Simpson from Arete Research.

I wanted to ask about your relationship with Huawei and how you see the impacts of the U.S. regulation on your business with Huawei in the second half of the year.

My understanding is that you will still have a relationship, you will still be shipping wafers probably at elevated levels in Q3.

But can you confirm whether or not you have any sales with Huawei in Q4?

And if not, how do you manage your 5-nanometer utilization given the importance of Huawei as a customer?

Okay.

Let me summarize your question, Brett.

It is regard to the relationship with Huawei but wants to know what is the impact on our business from Huawei in the second half of this year.

Will we continue to ship wafers to this customer in the fourth quarter?

If we do not, then how will we manage the impact to our 5-nanometer?

Okay.

Let me answer your question.

As C.C. just reported, we are complying fully with all the regulation.

And we did not take any new orders or production starts from this customer since May 15.

Although this regulation is just finished their public comment period, the BIS has not did a final ruling change at this point.

And so it's very early, still early to confirm, but under this current status, we do not plan to ship wafers after September 14.

And yes, there will be a challenge to work dynamically with other customers.

That's currently we are working with them.

And -- but as you heard, we made a -- C.C. just made of -- our 2020's guidance is above 20%.

That tells you we are relatively progressing well in filling up the left -- capacity left open.

Okay.

Brett, thank you.

Do you have a second question?

Yes, just a follow-up.

I wanted to ask about depreciation for this year.

I think previously you talked about mid- to high-teen growth of depreciation in 2020.

Can you confirm whether that's still the case?

And I look at the first half depreciation, and it looks like depreciation costs were down year-on-year.

And so in order to get to the mid- to high-teens growth, that would imply a large increase in depreciation sort of in the third and the fourth quarter.

So if you can just clarify exactly how we should think about depreciation for the next couple of quarters, that would be great.

Okay.

Brett is asking.

His second question is our depreciation outlook for 2020.

Do we still maintain what is our depreciation for 2020 year-on-year?

And then does this imply a pickup in depreciation in the second half on a quarterly basis?

Okay.

Brett, our current estimate on 2020 depreciation year-on-year growth is still high-teens growth.

So that gives you an idea of what the second half depreciation will be.

It will be higher than the first half.

Okay.

Thank you, Brett.

The next one on the line is Mehdi Hosseini from SIG.

I want to go back to your N4 and N3.

And how should we think about the migration and specifically to what extent this is driven by converting rather than installing new equipment?

And I have a follow-up.

Okay.

Sorry, Mehdi.

Let me make sure we understood your first question.

You're asking about N4 and N3, how to think about the migration.

And is there a conversion, tool conversion, involved between N4 and N3?

Is that your question?

Correct.

Okay.

All right.

Actually, the N4 is a kind of improvement, continuous improvement, from N5.

So it has a new -- improved the speed, improved the geometry just a little bit.

N3 is totally a new node, all right?

So that's N4 using the same equipment as N5.

N3, we expect it to have a high percentage of the tool continue to be used from the N5, but N3 is a totally new node.

Okay, thank you.

Do you have a second question, Mehdi?

Yes.

And my second question has to do with your HPC revenue growth in Q2.

It was significantly higher compared to Q1.

Will there be -- can you please elaborate which specific subsegment within HPC is doing better?

Is it driven by Communication or Computer?

And how do you see those trends trending into Q3?

So Mehdi, your second question is looking at our HPC sequential growth in the second quarter.

Mehdi wants to know what specific segments are driving that increase.

And what is the outlook?

Well, Mehdi, I don't think we want to break down the details on the different platforms.

Sorry about that.

Okay, Mehdi?

The concern is that maybe perhaps Huawei may have pulled in before you stop taking orders.

I'm trying to better understand how that particular customer has procured wafer in the first half versus second half.

Sorry, Mehdi.

We -- you know we don't comment on specific customer.

Okay.

Thank you, Mehdi.

The next one, we're having Randy Abrams from Crédit Suisse.

My first question.

I wanted to ask a bit more on the CapEx rate, as that's more a function of what you mentioned in the forward demand outlook.

If you could give a view on '21, I know it's in early stage, but just factoring a full year, where you mentioned Huawei and also mentioned potential -- but you don't rule out an inventory correction.

And it does seem like Samsung at least is discussing a bit about some graphics and high-end smartphone business.

So I'm curious, I guess, with the CapEx raise.

What's driving it?

If there's certain drivers that maybe lifted on the 2021, how you're seeing that.

And implications -- it follows up on Bill's question, but implication for 2021, if it seems like it might be a bit lower CapEx if you're spending a bit ahead of that now.

Okay.

So Randy, let me summarize your first question.

Your first question is really what is driving our raise for the 2020 CapEx.

What is the drivers for that?

And then what is the outlook 2021 CapEx?

CapEx and sales, the sales just factored in your comments about inventory if your competitor is taking a bit of business and also your view that we could have a -- or don't rule out an inventory adjustment.

The -- let me discuss.

CapEx is a -- we do the CapEx based on long-term perspective.

If you talk about -- this year's CapEx mainly, of course, this shows our demand of N5 is very strong.

And if you talk about the next year's CapEx, it's really talk about 2022's demand, which we see the continued increase of N5 demand.

And also, we see the starting the launch of N3 technology.

And we'll see by then how much the CapEx will increase, and we report to you in due time.

Okay.

Do you have a second question, Randy?

Yes.

And if I could follow up because you mentioned like so the higher CapEx for this year is a function that you expect next year to be even stronger.

So can you talk a bit about with -- I know you're talking about the megatrends, but I'm curious if you're thinking about just what you mentioned also, like could next year have impact from the high base this year on the inventory buildup and also the full year -- like in the first quarter, if Huawei is out, there's probably pent-up demand being tight, but how do you view a full year if you're not shipping to Huawei?

Unless you're counting on, by that point, some partial license to -- or if in the base case you're assuming not shipping to Huawei next year.

Okay, Randy's second question.

He wants to -- he is thinking that, with potential possibility of inventory correction with the U.S. regulations, will that impact?

What is the impact to 2021 growth outlook and CapEx?

Yes.

Randy, it's just too early for us to discuss anything about 2021, so we'll just wait until when the time approaches.

Okay.

Thank you, Randy.

The next one is Roland Shu from Citigroup.

First question is can you remind me again how does the inventory variation adjustment work every quarter.

How about with 3Q?

Is this inventory valuation adjustment favorable or unfavorable to the gross margin?

It's my first question.

And second question actually is you talk about that you are working with customers to minimize the impact of U.S. new regulation.

How are you going to do -- are working on that?

Okay.

So Roland, your 2 questions.

Your first question is what is the impact of inventory revaluation?

And then in the third quarter, will it be a favorable or unfavorable impact?

And your second question is you want to know how we are dynamically working with customers to mitigate the impact of the new U.S. regulation.

Correct.

Okay.

Let me -- Roland, let me make some comments on the inventory valuation adjustment first.

The impact on margins from inventory valuation adjustment is inversely correlated to that from changes in utilization.

We normally report the net impact on margins from these 2 factors together.

When we compare margins quarter-over-quarter, we will report the Q-on-Q change in impact from inventory valuation adjustments when it is more significant.

In the second quarter, the quarter-over-quarter change in impact from inventory valuation adjustments was more significant.

And if you ask about third quarter, at this moment, we believe the impact is less significant.

Okay.

And then...

Let me -- he's asking about how we work with customers dynamically to mitigate the impact of what we've done.

I cannot tell you that -- how we are going to do it because of this is our company's strategy and our strengths, but one thing I can tell you, we are based on the technology leadership and excellent manufacturing.

That's all we did.

Okay.

Thank you.

The next questions, Charlie Chan, Morgan Stanley.

So my first question is really about your upward revision of the full year revenue guidance.

So compared to last time, it was at mid-5% to 10%, and now it's above 20%.

I think there's at least 5 percentage points of revenue growth in 2020, but last time, your assumption is that pandemic can get controlled by June.

And now currently, there is a second-wave pandemics in many countries.

So how are you going to reconcile this kind of weak economy or health care crisis issue versus your very strong revenue guidance?

Should I just attribute that to the higher 5G smartphone penetration?

Or there's other factors that we should pay attention to.

Okay.

Let me summarize your first question, Charlie.

You're asking basically we have increased the full year outlook, but the risk of COVID-19 continues to remain.

So how to reconcile a weak global economy with TSMC's full year outlook.

And what peak -- will be driving this besides 5G smartphone preparation?

Well, Charlie, we do observe the 5Gs smartphone, the momentum is getting stronger, is -- so we understand the situation.

However, we also observe that our customer are making efforts to ensure supply chain security.

So bear in mind as there is a second wave, third wave of COVID-19, but it seems that end demand looks very promising.

So they are not afraid of -- to make sure that their supply chain will not be disrupted.

Because of 5G, as you just mentioned, 5G smartphones demand is continuing to increase.

Okay.

Do you have a second question, Charlie?

Yes, I do.

So I think a lot of things happened over the past months, right?

Another thing, I would take it as a U-turn, is your decision for the U.S. fab intentions because, half year ago, I remember, the comment was like the cost is pretty high.

Logistics doesn't make sense.

So what exactly is the trigger for you to change this U.S. operations decision?

And it would be very kind of you if I can add a very small question because the investors may care as well, your fourth quarter seasonality.

Because based on your new full year guidance, if we will take it at the 20% or 21% lower bar, the fourth quarter revenues may decline sequentially.

Is that a kind of fair comment?

Okay.

Well, Charlie, your second question relates to our U.S. fab plan.

And you want to know why, 6 months ago, we were talking about the cost gap being the major challenge.

And now we have decided to go ahead, so what has changed?

Yes, on that figure, yes.

Well, as you know, the -- with expanding our technology ecosystem and reach to global talent, closer to our customer to get a better service are all benefits of the fab in U.S., but in the past, indeed the cost, the gap, prohibited us to make those decisions.

More recently, I think since last December, and -- I think the things is getting a turn.

And we did get a positive encouragement from the U.S. administration and about the cost gap.

Then actually, the U.S. administration and the State of Arizona combined, they do -- they seem to be able to close the cost gap we used to hold up against this decision with their commitment, and we are preparing for that.

And how do they close the cost gap?

As you have been reading, we -- the U.S. congress, both in Senate and House, are all driving for the incentive packages aimed at revive U.S. semiconductor manufacturing.

And with that, I think they do have a way to fulfill that commitment to make up the cost gap.

And that was the major decision turning point.

And then okay.

And then, Charlie, he snuck in a third question, which he wants to know our outlook for fourth quarter given the full year guidance.

Okay.

Well, Charlie, it's also too early to talk about fourth quarter, but I think you can do the math and come up with certain estimation.

But what we can say is our second half will be growing, will be higher than the first half.

Okay.

Thank you.

The next one to ask questions, Bruce Lu, Goldman Sachs.

I think, given your positive progress in 3 nanometers and 5 nanometers and especially regional CapEx, can we assume that -- similar to previous node like 7-nanometer or 12 millimeters, that the first year of 3 nanometers can achieve 10% of the wafer revenue and second year of the 5-nanometer can achieve 30% of the wafer revenues?

Okay.

So Bruce, your first question is regards to N5 and N3.

Bruce wants to know what's the progress in N3.

Can it be -- contribute 10% of the wafer revenue in the first year?

And he also wants to know, can N5 contribute 30% of the wafer revenue in its second year?

Okay.

Bruce, both of them are really too early to talk about it.

We certainly hope that there will be pretty big nodes, but we will definitely let you know when time is closer.

Do you have a second question, Bruce?

Yes.

I think that -- I think just probably to check, we released our 5G penetration shipment or forecast, so we lowered the overall smartphone shipment forecast for 2020.

And how about the actual number for the 5G smartphone shipment?

Is that -- the penetration is up because of the lower total smartphone shipment?

Or the 5G smartphone shipment itself is going up as well.

Okay.

So your question, second question, Bruce, is that we -- the global smartphone shipment, we now lowered to low-teens decline, but we raised the 5G penetration to high teens.

Is this simply because of a lower -- smaller global base?

Or what is the 5G penetration number?

Well, the 5G penetration, as I said, momentum continued to increase.

So even with the total smartphone number being decreased, say, at the low teens, but 5G's percentage continued to increase.

And that's what we observe.

And also, the 5G's semiconductors content is higher than the 4G, and especially high end is much higher.

So that's what we based on.

Okay.

Thank you, Bruce.

The next on the line is Aaron Jeng from Nomura Securities.

Can I have a follow-up to Bruce's question just right now?

He was asking -- by lowering the total smartphone demand to low teens -- down 10% to 15% now from the earlier version of down 5% to 10% but raising the 5G penetration rate to 15% to 20% from earlier, only meeting, say, 15%.

And -- but in terms of absolute 5G phones demand or sell-in number, is the number being raised?

Or it's still pretty much the same as the prior versions.

That's a follow-up.

Actually this is the first -- a part of my first question, but it just happens to be a follow-up to Bruce's question.

Okay.

Aaron, let me summarize your first question.

Basically, Aaron wants to know with -- is the -- is our forecast for 5G smartphone in terms of units increased?

The answer is yes.

Okay.

What is your second question?

Okay, okay, let me -- so okay, let's ask my -- this question, that what -- and I was trying to compare the outlook offered by TSMC for the industry and the outlook given in the year beginning 6 months ago.

And in the beginning, TSMC was saying that semi excluding memory was going to grow by 8%.

And now it's going to be flattish to slightly grow, which means, I think, overall demand including everything is lower than it was 6 months ago.

The foundry growth in the year beginning was 17%, but now it's pretty much unchanged at mid- to high-teens growth.

TSMC's growth in the year beginning was above the industry growth.

Now it's above 20% growth.

Okay, so my question is, over the last 6 months, TSMC, along with actually everyone in the world, particularly in tech, has experienced 2 difficult challenges, including, one, COVID-19; and two, Huawei issues.

But it turns out that TSMC is doing even better than there was no 2 -- these issues.

And so I wonder -- actually, I think the CEO already answered that the 5G absolute unit demand is going to be higher than you saw 6 months ago, which is one, I think, a key reason, but it looks to me that the 2 factors -- 2 negative factors are still huge.

Actually, one -- either one of them is big, right, so -- but in terms of being greater than -- if there's no these 2 negative impacts.

So how do we think about this?

Earlier also, Chairman meant that...

Aaron, okay, I think -- let me summarize your question because it's quite long.

I think, to -- in essence, what you're asking is, when you look at the industry framework that TSMC provided in the beginning of the year and you look at the framework now, you point out that the semi ex memory growth in January, we said plus 8%.

Now we said flat to slightly up.

Foundry growth in January, we said 17% increase year-on-year.

Now we say mid- to high teens, but for TSMC, growth we're now saying greater than 20%.

So given the challenges, the -- in this year from COVID-19 and such, what is driving TSMC's stronger growth?

Well, I can answer that question by simply one word: technology leadership.

Actually, we see a very strong demand from our 7-nanometer and 5 nanometers technology.

And 5G -- again I would like to say that 5G's momentum is getting strong, including also HPC -- I'm sorry.

Okay -- it's all right, yes.

All right, thank you.

Next, we're having Gokul Hariharan, JPMorgan.

First of all, I just wanted to understand.

We are running at 20-plus-percent growth this year.

Any thoughts on -- I think we expect some of these negatives trends will pass.

Any thoughts on why we aren't changing our long-term 5% to 10% target, especially given you're also spending more CapEx?

So if ROIC is similar, then probably you need to be at a slightly higher growth rate.

That's my first question.

Second, I just wanted to understand what is management's view on how much of this year's outgrowth, compared with semiconductor industry, have been on some of those inventory builds that your customers have undertaken in terms -- over the last several years.

And there are very good years that TSMC outgrows the semiconductor industry or the foundry industry by a significant margin.

And this seems to be one of those years where even smartphone is not really growing.

It's actually declining, while TSMC is growing more than 20%.

So I just wanted to understand.

Quite a bit of that is because real demand and market share gain in leading edge, but any thoughts on how much of that do you feel is some of these inventory and supply chain security inventory that your customers are building?

Okay.

Gokul, let me summarize your 2 questions.

Maybe I'll start with the second question first.

You just want to know management's view, the fact that TSMC's growth in 2020 is outpacing the foundry industry.

Can we break down what is driving this?

How much of it is from supply chain efforts to ensure supply chain security?

How much of it is market share gains?

How much of it is due to leading edge?

We certainly -- at this time, I don't think we can separate them.

So clearly each one, that is because of technology, because of share gain, because of HPC or something like that.

Again I would like to emphasize the need on the leading-edge technology node N7, N5.

And that's why we again have our advantage.

Okay.

Okay.

And then your second question, Gokul, to repeat again, is that with the strong growth we see this year and the megatrends that we identify for the next several years, is there -- will there be a change in our long-term growth target?

Well, we continue to emphasize that we will be at the high end of 5% to 10% CAGR.

Remember this kind of a forecast is a rolling forecast, so we continue to have confidence then in our technology and also our market share and so our growth.

Okay.

Thank you, Gokul.

Next one, we are having Sebastian Hou from CLSA.

Sebastian, are you on the line?

Sorry.

I forgot to unmute.

Can you hear me now?

Yes, we can hear you.

Please go ahead.

Okay, okay.

So I have 2 follow-up.

First one is that the 5-nanometer revenue contribution is lower, from 10% to 8%, but total revenue outlook is plus 8%.

So if we do the math, then 5-nanometer revenue were probably lower by 15% compared to April.

And further compared to January guidance, it's actually 20% lower, so how do we attribute this?

Is it due to the customers that got sanctioned in May or any other reasons?

And furthermore, it also means that the other technology nodes are actually growing stronger, so I wonder what's driving the other applications in those.

And also, can you give us update on your expectation for growth for the 4 major platforms with the new revised-up guidance?

Okay.

So let me summarize Sebastian's question.

He wants to know the -- what is driving the difference in terms of N5 today versus 6 months ago and what other nodes then are stronger.

And then he also wants to know the 2020 outlook, growth outlook, by platform.

Okay.

Sebastian, actually compared to 6 months ago, our N5 revenue actually increases.

And so the other nodes, maybe you can double check the math, okay?

And then the 2020 growth outlook by the 4 platforms.

Okay.

All the platform will grow, except the automotive.

Okay?

Okay, okay.

My second question is also a follow on the 5G smartphone and also the total smartphone guidance you just gave and the other analysts asked about.

So one of the -- so it looks like the total 5G smartphone numbers -- absolute number is raised.

And I wonder, did you base the forecast on the final SoC numbers or based on the forecast that you're seeing from your smartphone SoC fabs.

Well, we base on the one we work with our customer.

So that's a number that the customer demand to TSMC.

Of course, they are also doing their forecasts as we did.

So all of that...

Okay.

Thank you, Sebastian.

Thanks, Sebastian.

Sorry.

Next one, we're having Mehdi Hosseini, SIG.

I'm a little bit confused and I was wondering if you can help me.

All the 5G smartphone data points suggest that the smartphones that are selling-through are priced worth less than USD 300.

And also, your commentary suggests that, despite the fact that COVID has had a second wave, your outlook is actually stronger, so how can I reconcile a 5G smartphone which is mostly driven by low end and the second wave of COVID with your outlook?

It's -- okay, Mehdi, your question is how -- your observation is that 5G smartphone sell-through is mainly coming through at the low-end 5G smartphones priced at USD 300 or less.

And with the potential second wave of COVID, how can you reconcile this low-end demand with what TSMC is seeing?

Is that correct?

And also your outlook for the year.

Because early -- in last earning conference call, you said your outlook is based on COVID node pricing by June, but it seems like there's a second wave.

So Mehdi is also asking because, in April, we said our outlook was premised on stabilization of COVID by June, but now it looks like COVID-19 continues.

We don't -- actually, we don't confirm there is a second wave of COVID-19 per se, but we did that one alone.

We do observe that our customers have demand to TSMC.

And you mentioned that 5Gs is only in the low end.

We do expect there is a lot of new 5G phones and pretty high end in the second half of 2020, and that's what we based on our assumption.

Okay.

Thank you...

May I have one follow-up on CapEx?

Okay.

Just the $1 billion of the increase to 2020 CapEx, is that equally distributed between front-end equipment and back end?

Or is it more in one particular area?

What's driving the incremental increase?

Okay.

So Mehdi, your second question is what's the increase in the CapEx guidance?

Is it more driven by the front end or the back end for 2020?

Yes.

Well, basically, it's front end.

Okay.

Thank you.

The next one, we are having Laura Chen from KGI.

Congratulations for the good results.

Actually, my question is also related to the advanced packaging.

I recall that we mentioned that we have about USD 3 billion for the advanced packaging last year for the revenue contribution.

I'm just wondering what's the latest guidance for this year.

Any revenue target for advanced packaging?

And also what are the plans you're looking for in this space?

On the incremental increased CapEx, do we also have some plan in this space?

Okay.

So Laura's question is related to the advanced packaging.

She wants to know.

Last year -- I -- it was not $3 billion.

It was $2.85 billion.

So what is the growth outlook for this year, number one?

And then what's the plan for advanced packaging, the outlook going forward?

Okay.

We expect that the advanced packaging will grow probably similar to our corporate average this year.

As to the CapEx increase, yes, a little bit but mostly at the front end and with the advanced technology.

Okay.

Do you have a second question, Laura?

Yes, yes.

My second question is about the legacy process and also like 28-nanometer.

We all know that advanced packaging, we are very strong and fully loaded.

I'm just wondering that -- for the legacy capacity and utilization rate and especially for 28-nanometer, as C.C. also mentioned before, that you see structurally overcapacity in this space.

Looking forward, can we expect improvement in second half or next year given our good progress in the RFIC or the CIS, et cetera?

Okay.

So let me summarize.

Your second question, Laura, is looking at our mature nodes, what is the utilization outlook for our mature nodes and specifically for 28-nanometer?

Do we see improvement in second half or 2021?

All right, let me answer that, our mature node or what we call specialty, our mature nodes are loading actually is quite good, except 28-nanometer, okay?

I still want to emphasize that 28-nanometer has been overcapacity for the whole industry, but we continue to improve it and slowly.

And of course, we can see, the CMOS image sensor and also other applications, that will move into 28-nanometer, but it's slower than we thought.

However, it will be improved.

We have confidence to say that.

Okay.

Thank you, Laura.

Next one to ask questions, Randy Abrams, Crédit Suisse.

Okay, yes.

First one, I wanted to just go back with a clarification on Huawei, if you're factoring in sort of the future view and these potential shipments.

I think one is the regulations seem to allow some ways to ship to Huawei.

I know you comply by the rules, but it seems to allow some way to shift directly to OSATs.

So I'm curious either from that or a perspective that you get a partial or a full license, if you're building that into the base case.

Actually, the current regulation spells do not prohibit it, standard product or general product, to be able to ship to Huawei.

And therefore, we think Huawei as a smartphone business will -- most likely, they may strategize to stay by procuring general-purpose products.

I think I -- Randy, part of your question is that, from TSMC's perspective, are there alternative ways to ship to this customer such as shipping to OSATs?

Or will we have a partial license?

No, no.

We don't have alternative way to ship.

Okay?

Okay, that's -- okay, if I can get the second question, if you could...

Randy, did you -- are you still there?

(inaudible) nanometer, if that's going to be the steep ramp-up, will be available in 2020.

Sorry.

Randy, you dropped off for a second.

Can you repeat your second question again?

Yes.

It's actually more about the half-nodes.

The 4-nanometer will be available.

I think mass production early 2022.

So with 3 coming out late in the year, if you're expecting that would be the steep ramp so we could see high volume.

It also could allow you, I think, with the tool reuse, a bit lower spend.

So I'm curious how you're thinking about that.

And then also on the 6-nanometer, if you're still seeing most of the customers on 7 migrate to 6, where I think previously you expected majority could end up going to that half-node.

Okay.

Your second question, Randy, is related to N4 and N3.

And thus, were we -- with the timing differences of N4 and N3, will we see a lower spend as a result?

Randy's view is that N4 will be early 2022.

N3 will be late 2022.

And then so with some conversion, will that result in lower spend?

And he also wants to know for our N6.

We've talked about it before.

Do we see still a strong migration of our customers from N7 to the ex-N6?

Well, let me answer the second one first.

On the N6, yes, we have been -- offer to our customer with a compatible -- actually, it's fully compatible to N7, so it will have a very good opportunity to catch the second wave of 7 nanometers product.

With the same kind of strategy, we offer N4 to follow the N5.

So we do expect that N5 is a product -- finally, a lot of -- a large portion of the N5 product will move to N4.

So it's not the -- through mix that we say N3 is a progress or N3 is a ramp-up.

N3 is another full node.

It's not -- it's more advanced by nature as -- than N5.

So N3 is N3.

N4 is N4.

And then N6, do we still see strong...

Yes, we -- I already said that N6 is following the N7.

Sorry.

Okay, okay.

Next one, we're having Charlie Chan, Morgan Stanley.

So 2 parts.

First is about your N3 and CapEx.

Do you spend some CapEx for N3 this year so this is another reason why you see a CapEx upwards revision?

Okay.

Charlie, your first question is that, for 2020 CapEx, do we spend -- does it include spending for N3?

Charlie, part of the CapEx this year is for N3, but that's not the reason for our increase in CapEx.

Okay.

And do you have a second question, Charlie?

Yes, I do.

So every quarter, I ask this question about the Chinese competition.

And notably, your China competitors make very few IPO appeals with a very high valuation.

And supposedly, those few companies can spend for their future CapEx or even revenue growth, right?

So I'm not saying about in the recent quarter, but in the long term, do you think that there is a threat and you probably may lose market share to China players given they want localization?

Or do you have any China strategy to accommodate to China's localization policy?

Okay.

So Charlie, your question is that what is the threat from Chinese foundry competition?

Do we see it as a growing threat?

How do we respond?

Yes.

Well, Charlie, I also answer every time that we compete in technology and the manufacturing and the customer relationship.

And whether it's in China, in other area, we stay the same.

We're competing in technology, manufacturing.

And we have been keeping very good relationship with our customer.

We've won their trust.

Okay, thank you.

In the interest of time, I think we will take the last 2 callers on the line.

Next one, we are having Bruce Lu, Goldman Sachs.

The first question is for the norm for the capital intensity.

I think remember -- I remember, like 6 months ago, the management was talking about capital intensities will go back to 30% to 35% for 2021, but earlier, management was talking about what -- you will go back to closer to 35%.

So do we foresee that the capital intensity norm will be closer to 35%?

Or the norm will still remain at 30% to 35%.

Okay.

Bruce, I think our comment is, over the long run, it will go to about 35%, and that remains the same.

Okay.

I understand that.

Second question is that we saw that TSMC announced like 2 new factories for the packaging this year, just did the groundbreaking, and the size for the factory is pretty big.

Do we anticipate that the advanced packaging penetration rate will be a lot higher in the advanced node?

And what's the future outlook for the advanced packaging?

Okay.

So your second question is the -- Bruce wants to know that we announced a large advanced packaging site recently, so he wants to know what is the penetration rate, so to speak, for advanced packaging in the leading nodes going forward.

And what is the outlook?

Well, we do work with our customer closely.

And we do see some increase on the demand of advanced packaging, and therefore we try to enlarge our capacity.

That's for sure, but let me stress that we enlarge our advanced packaging's capacity.

It's for leading edge, also for specialties.

There is a new demand coming out and we have to work with our customer to meet their requirement.

Okay.

Thank you, Bruce.

The last one to ask questions, Sebastian Hou, CLSA.

So a follow-up on the CapEx increase, the $1 billion CapEx increase.

Which particularly nodes do that go to?

Sebastian wants to know, with the increase in the CapEx to $16 billion to $17 billion from $15 billion to $16 billion previously, what node is the CapEx spending going to?

Leading edge...

It's leading edge, leading edge...

Okay, got it.

Then maybe last question is I think the U.S. senator has proposed 2 bills, CHIPS Act and American Foundries Act, in June.

So I wondered if the -- how does that roll with TSMC Arizona plant?

And if that were to be passed, that bill will -- should be passed, whether TSMC or a non-American company are eligible for full for potential subsidy.

Well, I...

Sebastian, just to make sure we understand your question.

Your question is related to some of the proposed regulations in the U.S. such as the CHIPS Act and the AFA.

If these bills were to be passed, would it be eligible for TSMC or the industry?

Yes, it's well aligned with our request.

And if those bills in different forms passed, I think the administration and State of Arizona will make this project happen.

Okay.

Thank you, Sebastian.

Thank you.

This concludes our Q&A session.

Before we conclude today's conference, please be advised that the replay of the conference will be accessible within 4 hours from now.

The transcript will be available 24 hours from now, and both of them will be available through TSMC's website at www.tsmc.com.

Thank you for joining us today.

We hope everyone continues to stay healthy and safe.

And we hope you will join us again next quarter.

Goodbye, and have a good day.