大家午安。我是台積電法人關係處的蘇智凱。歡迎您參加台積公司2022年第三季的法人說明會。由於本法說會是向全球投資人同時連線轉播，所以我們會全程使用英文，請您見諒。Good afternoon, everyone, and welcome to TSMC's third quarter 2022 earnings conference call. This is Jeff Su, TSMC's Director of Investor Relations and your host for today. 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. If you are joining us through the conference call, your dial-in lines are in listen-only mode. 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 third quarter 2022, followed by our guidance for the fourth quarter 2022.
Afterwards, Mr. Huang and TSMC CEO Dr. C.C. Wei will jointly provide the company's key messages. Then we will open the line for Q&A. 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. Now I would like to turn the call over to TSMC CFO Mr. Wendell Huang for the summary of operations and the current quarter guidance.
Thank you, Jeff. Good afternoon, everyone, and thank you for joining us today. My presentation will start with financial highlights for the third quarter 2022. After that, I will provide the guidance for the fourth quarter. The third quarter revenue increased 14.8% sequentially in NT dollar, or 11.4% in U.S. dollars, as our third quarter business was supported by strong demand for our industry-leading 5 nm technology. Third quarter gross margin increased 1.3 percentage points sequentially to 60.4%, slightly ahead of our guidance as we enjoyed a more favorable foreign exchange rate and cost improvement efforts. Total operating expenses accounted for 9.8% of net revenue as compared to 10% in the previous quarter.
Operating margin increased 1.5 percentage points sequentially to 50.6%, mainly due to better operating leverage. Overall, our third quarter EPS was NT 10.83, and ROE was 42.9%. Now, let's move on to revenue by technology. 5 nm process technology contributed 28% of wafer revenue in the third quarter, while 7 nm accounted for 26%. Advanced technologies, which are defined as 7 nm and below, accounted for 54% of wafer revenue. Moving on to revenue contribution by platform. Smartphone increased 25% quarter-over-quarter to account for 41% of our third quarter revenue. HPC increased 4% to account for 39%. IoT increased 33% to account for 10%. Automotive increased 15% to account for 5%. DCE decreased 2% to account for 2%.
Moving on to the balance sheet. We ended the third quarter with cash and marketable securities of NT 1.5 trillion. On the liability side, current liabilities decreased by NT 38 billion, mainly due to the decrease of NT 116 billion in short-term loans, partially offset by the increase of NT 70 billion in accrued liabilities and others. Long-term interest-bearing debt increased by NT 88 billion, mainly as we raised NT 60 billion of corporate bonds during the quarter. On financial ratios, accounts receivable turnover days decreased 1 day to 36 days. Inventory days decreased 5 days to 90 days, primarily due to higher wafer shipment during the quarter.
Now let me make a few comments on cash flow and CapEx. During the third quarter, we generated about NT 413 billion in cash from operations, spent NT 266 billion in CapEx, distributed NT 71 billion in fourth quarter 2021 cash dividends, and raised NT 60 billion from corporate bond issuances. Overall, our cash balance increased by NT 43 billion to NT 1.3 trillion at the end of the quarter. In U.S. dollar terms, our third quarter capital expenditures totaled $8.75 billion. I have finished my financial summary. Now let's turn to our current quarter guidance. Based on the current business outlook, we expect our fourth quarter revenue to be between $19.9 billion and $20.7 billion, which represents a point.
A 0.4% sequential increase at the midpoint. Based on the exchange rate assumption of $1 to NT 31.5 , gross margin is expected to be between 59.5% and 61.5%. Operating margin between 49% and 51%. This concludes my financial presentation. Now, let me turn to our key messages. I will start by making some comments on our third quarter and fourth quarter profitability. Compared to the second quarter, our third quarter gross margin increased by 130 basis points sequentially to 60.4%, mainly due to a more favorable foreign exchange rate and cost improvement efforts despite continued inflationary cost pressures.
Compared to our third quarter guidance, our actual gross margin exceeded the high end of the range provided three months ago as our guidance was based on exchange rate assumption of $1 to NT 29.7 . Whereas the actual third quarter exchange rate was $1 to NT 30.32 . This created about 80 basis points difference in our actual third quarter gross margin versus our original guidance. We have just guided our fourth quarter gross margin to be flattish sequentially to 60.5% at the midpoint, as a more favorable exchange rate assumption will be offset by a lower capacity utilization rate. As a reminder, six factors determine TSMC's profitability, leadership, technology development, and ramp-up, pricing, cost reduction, capacity utilization, technology mix, and foreign exchange rate.
Looking ahead to 2023, we face challenges from N3 ramp dilution, higher year-over-year increase in depreciation costs, rising inflationary costs, semiconductor cyclicality, and overseas fab expansions. To manage our profitability in 2023, we are working closely with our customers to support their growth and continue to strategically and consistently sell our value. We are also working diligently on our internal cost improvement. Excluding the impact of foreign exchange rate, of which we have no control over, and taking the other five factors into consideration, we believe a long-term gross margin of 53% and higher is achievable. Next, let me talk about our 2022 CapEx. As I have stated before, every year, our CapEx is spent in anticipation of the growth that will follow in future years.
Three months ago, we said our 2022 CapEx would be closer to the lower end of our $40 billion-$44 billion range. Now, we are further tightening up this year's capital spending and expect our 2022 CapEx to be around $36 billion. About half of the change is due to capacity optimization based on the current medium-term outlook, and the other half is due to continued tool delivery challenges. Out of the around $36 billion CapEx for 2022, between 70%-80% of the capital budget will be allocated for advanced process technologies. About 10% will be spent for advanced packaging and mask making, and 10%-20% will be spent for specialty technologies. Looking ahead, we will continue to manage our business prudently given the near-term uncertainties, and adjust and tighten up our capital spending where appropriate.
That said, our commitment to support customers' growth remains unchanged, and our disciplined CapEx and capacity planning remains based on the long-term structural market demand profile. We will continue to work closely with our customers to plan our long-term capacity and invest in leading-edge and specialty technologies to support their growth while delivering profitable growth to our shareholders. Now, let me turn the microphone over to C.C.
Good afternoon, everyone. First, let me start with TSMC's more resilient near-term demand outlook. We concluded our third quarter with revenue of NT 613.1 billion, or $20.2 billion, supported by strong demand for our industry-leading 5 nm technologies. Moving into fourth quarter 2022, we expect our business to be flattish as customers' ongoing inventory adjustment is balanced by continued ramp-up of our 5nm technologies, supported by smartphone and HPC applications. We expect our full-year growth in 2022 to be mid-30% in U.S. dollar terms. On the demand side, we continue to observe softness in consumer end-market segment. Other end-market segments, such as data center and automotive-related, remain steady for now for TSMC. We start to see the possibility of adjustment down the road.
On the inventory side, our customers and the supply chain continue to take action to adjust their inventory. We expect the semiconductor supply chain inventory to peak in third quarter this year and start to reduce in fourth quarter this year. We also expect it will take a few quarters through first half 2023 to rebalance to a healthier level. While the ongoing inventory correction will also affect TSMC, we expect our business to be less volatile and more resilient than the overall semiconductor industry during this period, supported by three key factors that are TSMC's core strengths in the foundry industry. First, our technology leadership and differentiation is much stronger today as compared to previous years. This enable TSMC to win business and enables our customer to win business in the end markets despite the semiconductor cyclicality.
Secondly, through our comprehensive design ecosystem and optimize the process technologies, we are able to address and capture the structural increase in demand for computation and build a strong portfolio in high performance computing. Third, our strategic relationship with our customers are long-term in nature, and we continue to work closely with our customer on technology development, capacity planning, and pricing to support their long-term demand and growth. As a result, we continue to see strong demand for our leading node, except N7 and steady demand for our differentiated specialty technologies and mature node. Looking ahead to 2023, with the successful ramp-up of N5, N4P, N4X, and the upcoming ramp of N3, we will continue to expand our customer product portfolio and increase our addressable market.
Thus, while the ongoing semiconductor inventory correction will affect our first half 2023 utilization rate, we expect our business to be supported by stronger demand for our differentiated and leading advanced and specialty technologies, and for 2023 to be a growth year for TSMC. Next, let me talk about the N7, N6 demand outlook. Due to end market weakness in smartphones, PCs, and customer product schedule delays, starting 4Q this year, our N7, N6 capacity utilization will not be as high as it has been in the past three years. We expect this to persist for a few quarters through first half 2023 as the semiconductor supply chain inventory takes a few quarters to rebalance to a healthier level, and we have adjusted our N7, N6 CapEx accordingly.
We believe the N7, N6 demand is more a cyclical issue rather than structural, and expect our N7, N6 demand to pick up in second half 2023. Longer term, we continue to work closely with our customers to develop specialty and differentiated technology, and are confident to drive additional wave of structural demand to backfill our N7, N6 capacity over the next several years. Thus, our 7 nm family will continue to be a large and long-lasting node for TSMC. Now, I will talk about our N3 and N3E status. Our N3 is on track for volume production later this quarter with good yield. We expect a smooth ramp in 2023, driven by both HPC and smartphone applications. Our customers' demand for N3 exceeds our ability to supply, partially due to the ongoing tool delivery issues. We expect N3 to be fully utilized in 2023.
We expect N3 revenue in 2023 to be higher than N5 revenue in its first year in 2020, and for N3 to contribute a mid-single-digit % of our wafer revenue in 2023, as our overall revenue base is much larger today than in 2020. N3E will further extend our N3 family with enhanced performance, power, and yield, and offer complete platform support for both smartphone and HPC applications. N3E development is progressing ahead of plan, and volume production is now scheduled for second half 2023. Despite the ongoing inventory correction, we observe a high level of customer engagement at both N3 and N3E, with a number of tape-outs more than 2x than that of N5 in its first and second year.
Thus, we are working closely with our tool supplier to address the tool delivery challenges and prepare more 3 nm capacity to support our customers' strong demand in 2023, 2024, and beyond. Our 3 nm technology will be the most advanced semiconductor technology in both PPA and transistor technology when it is introduced. We are confident that N3 family will be another large and long-lasting node for TSMC. Finally, let me talk about the future driver of leading node adoption. TSMC's mission is to be the trusted technology and capacity provider for the global logic IC industry for years to come. Our job is to help our customers unleash their innovations and enable them to capture greater value and win in their end markets.
As the industry continues to pursue scaling, it is true that geometry shrink is slowing down and becoming more challenging for everyone due to rising process complexity. However, it is also true that demand for energy-efficient computing is accelerating in an intelligent and connected world as technology is becoming more pervasive and essential in people's life. Thus, the semiconductor industry value in the supply chain is increasing, and the value of technology platform is expanding beyond the scope of geometry shrink alone and increasingly toward greater power efficiency. As a result, our customers value much more than simply transistor cost. System performance and power efficiency has become key motivation for customers to adopt our leading node technologies. By working closely with our customer on technology development, our N3 and N2 will deliver full node strides in performance and power benefits while offering the industry's most advanced transistor scaling.
We expect strong demand for our leading node technologies driven by both smartphone and HPC applications to fuel our long-term revenue growth of 15%-20% CAGR over the next several years in U.S. dollar terms. With our leadership in both leading-edge process technology and 3D solutions, TSMC's technology cadence remains constant to deliver the value of our technology platform. We will continue to extend our overall competitiveness and technology leadership while delivering a predictable technology cadence that helps our customers to enhance their product competitiveness and grow their markets well into the future. This concludes our key message, and thank you for your attention.
Thank you, C.C. This concludes our prepared statements. Before we begin the Q&A session, I would like to remind everybody to please limit your questions to two at a time to allow all the participants an opportunity to ask questions. Should you wish to raise your question in Chinese, I will translate it to English before our management answers your question. For those of you on the call, if you would like to ask a question, please press the zero then one on your telephone keypad now. If at any time you would like to remove yourself from the questioning queue, please press zero two. Now let's begin the Q&A session. Operator, can we please proceed with the first participant on the line?
Yes, Jeff. The first one to ask question, Gokul Hariharan from JPMorgan. Go ahead, please.
Yeah, good afternoon, and thanks for taking my question. Congrats on the great results, especially in the margins. First question is on N7 and N6, this utilization slack that we are observing. Could we give a little bit more detail on why that is happening and why we think this is a short-lived couple of quarters issue, and we get a big pickup in the utilization in second half of the year? I think last time we saw this was for 20 nm, but that lasted for much longer period of time. Could you also give us some kind of comparison with what happened back in 20 nm and why this is going to be very different?
Maybe a little bit more color on what are the areas of backfill demand for N7 and N6 as the high-end smartphone processors and HPC start to move on to N5 and then N3. That's my first question.
Okay. Gokul, please allow me to summarize your first question. Gokul's first question is on N7, N6. He wants to understand what is driving the utilization slack, why we believe it is short-lived, and that the demand for N7, N6 can pick up in second half. Also on the little bit longer term outlook, why we believe that N7 can be backfilled, and it will not be like N28 a few years ago with a few years of underutilization.
Well, Gokul, let me answer your question first on the why demand drop. Demand drop because of market become soft, as we said in the weakness in the smartphone and PCs. Also a big factor is our customers' product schedule delay. All in all, put all together, we think that our utilization has been affected in first quarter this year and all the way to the first half of 2023. For the longer term, we continue to work closely with our customer. Actually, let me also say that this is a cyclical issue, so it will pick up anyway. We believe it will pick up in the second half of 2023.
For the longer term, we continue to work closely with our customer to develop specialty and differentiated technology to drive additional wave of structural demand from consumer, RF, connectivity, et cetera, and other application to backfill our N7, N6 capacity, and so for the next several years. That's all I
Does that answer your first question, Gokul? Gokul?
Yeah. I'm here. Sorry.
Okay.
My second question would be on the CapEx outlook. Now that we bring down the CapEx slightly for 2022, could you talk a little bit about what is the outlook direction for 2023 at least, if not absolute numbers, but just direction? Are we going to be flattish or it's likely to be going down? Second is also, could you also give us an update on the schedule for the new fabs, Kaohsiung, Kumamoto, Nanjing, and Arizona? Is there any change in the schedules in terms of these fabs coming into production? Thank you.
Okay. Thank you, Gokul. Gokul's second question is on CapEx and capacity. First he wants to hear, of course, you said that we have tightened up our CapEx in 2022, so he is asking for 2023 directionally, is there an indication of 2023 CapEx? And then he would also like an update on our expansion plans in Kaohsiung, Nanjing, Arizona, and Kumamoto.
Okay. Gokul, let me answer you the first part. For 2023 CapEx, it is too early to comment. We will provide you with a specific guidance in January. As we said, we have tightened up our 2022 CapEx to reflect the current macro outlook, as well as to delivery issues. Looking ahead to 2023, we will continue to be careful and manage our business prudently given the near term uncertainties. We will adjust and tighten up our capital spending where appropriate, but we will continue to work closely with our customers to invest for the long-term structural market demand profile to support their growth. Well, let me answer the second part of Gokul's question. Gokul, you ask about the progress of our Arizona fab, Nanjing fab, and Kaohsiung.
Let me say that Arizona fab will continue and on schedule. There's no doubt about it because this is an N5 family. We still have a very strong demand. For Nanjing, we just get our one-year authorization for 28 nm expansion, so it is on schedule also. For Kaohsiung, initially we planned two fabs at the beginning of a 28 nm expansion and the N7. Now, N7 has been adjusted and 28 nm expansion is continued and on schedule.
Also Kumamoto?
Oh, okay. Also, the Japan fab is on schedule. To meet the customers' demand. Okay. Gokul?
Got it. Thank you very much.
All right. Thank you, Gokul. Operator, can we move on to the next participant, please?
The next one to ask questions is Bruce from Goldman Sachs. Go ahead, please.
Thank you for taking my question. My first question is regarding to the HPC, which is a key growth driver for TSMC for the coming years. However, with the recent U.S. new restriction to China, what do you think about the HPC demand moving forward? What kind of impact? Is going to see a slowdown from China, or we're gonna see the acceleration from the non-China side?
Okay. Bruce's first question is on HPC. He notes HPC, we have said repeatedly, will be TSMC's key growth driver and main engine the next few years. He wants to know, I believe, Bruce, the impact of the recent U.S. regulations, does that affect the overall.
HPC demand, or the overall profile. Is that correct?
Yes. What's the impact from this new restriction to TSMC and overall industry?
Okay. Let me answer that. Bruce, it's based on our initial reading and feedback from our customers. The new regulation set the control threshold at very high-end specification, which is primarily used for AI or supercomputing applications. Therefore, our initial assessment is the impact to TSMC is limited and manageable. We are continuing to closely monitor the situation to ensure that we are all in full compliance with all the rules and regulations. For the longer term, it's too early to really assess all the true impact or influence, but we will give you the update in the following earnings call.
Okay. Okay, understand that. Thank you. My next question is regarding to the cyclical nature for the 7 nm. We also noticed that most of your other nodes, you know, the capacity utilization rate is still at a very high level or at least much better than 7 nm as management mentioned. Why is 7 nm so cyclical? Because, you know, maybe because you guys are too big that you have to keep up with the industry or what is the difference between your 7 nm and your other node?
Okay, Bruce's second question is looking at N7 specifically. He wants to know, you know, utilization still seems to be holding up well. Why specifically N7 is more cyclical and utilization is, you know, not as high as it has been. Is that correct, Bruce?
Yes.
It just happened, you know, that most of my smartphone and PCs customer are using N7, N6 node. You know, it just happened. The market weakness in the smartphone and PC happened at the same time. Also other customers or product schedule delay. All in all, that's why it become lower utilization rate as compared with other node. If you want to compare with the N5 or you compare with the 28 nm, all those are very, you know, is still a very high demand, and we continue to enjoy the higher market share.
I want to dig in a bit because we get used to like, you know, TSMC will manage your customers' product and overall outlook, even with some paperwork delay, you can take it up with someone else. You know, anything different with this time that 7 nm you cannot have as good as, you know, your other node.
Bruce wants to still understand, you know, why the 7 nm utilization cannot be as high as the other nodes if, you know, we work closely with customers and it's well planned. Well, again, Bruce, we work closely with our customers, but our customers got caught in this inventory correction and the market downturn. They didn't know this one probably two quarters before. You know, at the beginning of this year, they all still give us a very high number of their forecast. You know, it just happened. As we said, we believe this is a cyclical issue, and it will pick up. Before that, it probably will take a few quarters.
I understand. Thank you. We just get too used to it to be like, to expect TSMC always deliver a much better result.
I understand.
Thank you.
Okay. Thank you, Bruce. Operator, can we move on to the next caller, please?
Next we have Randy Abrams from Credit Suisse.
Yes. Okay, thanks for taking my question. Yeah, I wanted to ask a two-part first question. You mentioned in the prepared remarks about HPC and auto continue to be stable, but it seemed the signal may change. If you could give your view how you see inventory levels and some of the forward demand outlook from those areas. The second part, just wanna see if we could get a bit better visibility on first half. First quarter, how do you see it versus normal seasonal? For the trough, do you expect first quarter could be the bottom, or do you see the trend that we could be towards second quarter?
Okay. Randy's question is really looking at again from the end demand segment, data center and automotive, we say are remaining steady at TSMC for now. He wants to understand, though, what is the outlook down the road and what does that mean for inventory levels. Then secondly is also he wants to see if we can provide some more granularity about first quarter outlook versus seasonality, and whether we think the first quarter can be the bottom for the industry. Is that correct, Randy?
Yeah, that's correct.
Okay. Let me answer this one. Of course, we say that, so far our customer give us their demand forecast, so the data center and automotive related are still steady. Now the market becomes soft, and we are taking more conservative way in our planning for 2023. That's why we say that, we don't rule out the possibility there might have some correction also, but, you know, we did not see it right now, to be frank with you. For the inventory correction in the 2023, all we want to say is like that, we expect probably 2023, the semiconductor industry were likely to decline.
TSMC also is not immune, but we believe our technology position, strong portfolio in HPC and longer term strategic relationship with customer will enable our business to be more resilient than the overall semiconductor industry. That's why we say 2023 is still a growth year for TSMC, and the overall industry will probably decline.
Okay. I'll ask a quick follow-up and then the second question. I guess just the sense because it's still very firm in fourth quarter, if you're seeing a pretty meaningful fall off into first quarter, like some years 10%, 2019 declined over 20%. Just try to get a rough feel the type of decline factor when you're gaining some content and share. That's the follow-up to the first. Then the second question, actually just on gross margin because of a lot of headwinds I think Wendell will be talked about. If you could give just on a couple of these, depreciation factoring the recent CapEx, how much up and whether it's also front end or more back half loaded, given the 3 nm ramp.
For the startup cost for N3, if you expect similar to N5, given smaller percent of revenue, how much dilution from N3? Okay. Thank you.
Okay. Randy. Let me summarize. Randy, quick, I think we will not comment on the first quarter, but I think we can make some comment just sort of in terms of the overall inventory picture, in the, you know, looking into next year.
Well, actually, if you ask my opinion on the inventory picture and the inventory correction, it's too early to provide a specific number. However, the inventory correction will likely see its biggest impact sometime in the first half 2023.
The second part related to the gross margin, maybe Wendell can address what is the outlook for depreciation next year. Is it front or back-end loaded? What type of dilution we expect from 3 nm as it ramps in 2023.
Okay, Randy Abrams. The full year depreciation for next year is too early to talk about that. This year is mid-single-digit increase year-on-year. Next year, we expect it will likely to be meaningfully higher. We will give you the guidance in January. As to the dilution from N3, it will be between two-three percentage points from the whole year basis of our gross margins.
Okay. Actually one follow-up. Do you think on a single quarter with the inventory correction, do you expect to keep the 53% and above, factoring we're coming from a very high level, even through that first half inventory correction?
Right. It is too early to talk about 2023 gross margin, including a quarter over quarter. Even with all these cost challenges, we believe our structural profitability can be maintained, and we are confident to deliver a long-term gross margin of 53% and higher. Okay?
Okay. Great. Thanks for all the help.
Thank you, Randy. Operator, can we move on to the next participant, please?
Next one to ask questions, Charlie Chan from Morgan Stanley.
Hi. Good afternoon. Thanks for taking my question. My first question to management is about whether the company consider a share buyback, because the company seems to suggest the recent inventory correction is just a cyclical. You're still very positive on long term. I think our share price really reveal company's long-term value, right? Also, shareholders value. I'm not sure if a company want to do share buyback or cash returns.
Okay. Charlie's first question is, that, you know, although there is cyclicality, the long-term outlook appears good. Would the company consider doing a share buyback?
Yeah, Charlie, we constantly review all the different options of returning cash to shareholders. For share buyback at this moment, we are not considering it. We think our cash on hand will be better kept to invest in our capital expenditure to make a better return for our shareholders.
Okay. Understood. Thank you. Also, my second question is also about the U.S. sanctions impact. I know the company said that impact is limited and the long-term impact remains to be seen, right? My question is about the China market to TSMC. Is it as strategic as before after this event?
Okay. Charlie, second question, he wants to know with the recent U.S. regulations and the impact, how do we see the China market? Is it still a strategic market for TSMC? Is that correct, Charlie?
Yeah. How does company think about China for TSMC's kind of long-term picture? How important is China going forward?
Charlie, I would like to say that every region is important to TSMC. However, let me say that under the condition of work compliance with all the rules and regulations, TSMC will continue to serve all the customers all over the world. That's our position.
Okay. Including China.
All the customers.
Okay. Got it. Thank you. Those are my questions. Thank you.
Thank you, Charlie. Operator, can we move on to the next participant, please?
Next one to ask questions, Sunny Lin from UBS.
Hi. Thank you for taking my questions. Congrats on the steady performance. My first question is on the geopolitical tensions. I wonder, given some of the considerations regarding the geopolitical issues, how would you evaluate the longer term impact from customers potentially diversifying from TSMC foundries? And how are you managing the risk? Thank you.
Okay. Sunny's first question is that, given the geopolitical tensions, you know, how do we evaluate the long-term impact, and the risk of customers using other foundries? Is that correct, Sunny?
That's right. Thank you.
Okay. Sunny, we still believe the most important thing is the technology leadership in manufacturing and our customer trust. In different location, manufacturing or whatever, still, we still think that technology leadership is the most important thing. That's our strategy. We make it simple, that technology, manufacturing and customer trust.
Got it. A quick follow-up, is that going forward, should we assume an acceleration of your overseas expansions, just to diversify the production sites? i.e., if there could be a fab built in Europe?
Okay. Sunny's second question, can she assume that we will continue to increase the overseas global footprint expansion and also particularly in Europe? Well, we will continue to increase our overseas portion in manufacturing based on customers' need, in fact, based on the business opportunity and also based on the operations efficiency and economics. Whether we are going to be in Europe, we are in preliminary evaluation and do not rule out any possibility. Again, I would like to say the decision will based on customers' need, business opportunities, operation efficiency and cost economics.
Got it. Thank you. I actually have a second question, on HPC. With the increasing usage of chiplets, how would you manage the risk, in the case, that some dies are made at the other foundries and that have production issues and therefore impacting the production at TSMC as well? Thank you very much.
Okay. Sunny's second question is, with increasing usage and adoption of chiplets in HPC, how would we manage the risk in case, I think, Sunny, you're saying if dies at other companies or places have production issues, how do we, you know, manage the risk of that impacting TSMC? Well, Sunny.
That's right. Thank you.
Sunny, in fact, we would like our customer manufacturing every chip inside TSMC for sure. If there's a case that they have to use other companies or dies, we will work with our customer closely and minimize all the risk that as time goes by. That's what we are doing right now.
Got it. Thank you. That's very helpful. Thank you.
Thank you, Sunny. Operator, can we move on to the next participant, please?
Right now we have Laura Chen from Citigroup.
Hello. Hi. Thank you for taking my question. I appreciate that if you can share with your latest plan in your Nanjing fab. Like C.C. Wei already mentioned, you got the license for the 20 nm in Nanjing. I'm just wondering, do you also need the license for the 60 nm in the Nanjing fab?
Also, going forward, what's your plan of your operation in China? That's my first question. Thank you.
Okay. Laura Chen's first question is about Nanjing fab and our plans. She notes that we have received the one-year authorization, so our 28 nm expansion continues as planned. Her question is, do we also need a license for the 16 nm that we have in Nanjing? And then also her also what is our long-term future expansion plans in China?
Sunny, let me answer the first part.
Laura.
Sorry, it's Laura. Let me answer the first part.
Yes.
The one-year authorization that we received cover the Nanjing facility, so it's both the 28 and 16.
Okay, great. Thank you.
The second part is, what is our long-term expansion plans for China?
Okay. As C.C. said, we will be operating, serving all the customer under the condition that we will, fully follow and in compliance with all the rules and regulations.
Okay. That's clear. My second question is also about the fab globally in the longer term. We know that in overseas operation, usually they will have a much higher operational cost. How would that impact TSMC's long-term margin trend in our view? Maybe you can give us some estimation about your estimate of the percentage of margin in different region or the cost difference comparing to Taiwan.
Okay. Laura's second question is around our global expansion of our global manufacturing footprint. She notes she wants to know that overseas, fabs, are the costs higher? Do we have a breakdown how much the cost difference is in Japan and U.S. versus Taiwan? And then overall, with overseas expansion and if they are higher costs, how does this impact our long-term profitability and margin?
Okay. Laura-
Thank you. Yes.
Let me answer this question. The initial costs of overseas fabs are indeed higher than TSMC's fab in Taiwan, and it's mainly because of higher labor costs in different layers of the supply chain. We continue to work closely with the U.S. government as well as with our customers and supply chain partner to manage and minimize the cost gap. Now, through these efforts, we believe we can continue to earn the proper return and deliver the long-term growth margin of 53% or higher.
Okay, that's very clear. Thank you very much.
Thank you, Laura. Operator, can we move on to the next participant, please?
Next one to ask questions is Rolf Bulk from New Street Research.
Thank you for taking my question. I was hoping you could give some more context around the 3 nm tool shortages that you mentioned. Is that primarily lithography and then, specifically EUV related? Or do you also see shortages in other tool segments? Thank you.
Okay. Rolf's first question is around 3 nm and the tool shortages. He wants to know if this is very, just specific to lithography tools and EUV specifically, or is it more, I guess, broad-based.
Let me answer the question. Actually, it's a more broad-based because of our demand is high and certainly the photolithography tool is included, and one of the most important one.
Thank you. That's very useful. As my follow-up question, it would be great to get an update on your N2 nodes. With your current visibility, is the node still on track timing-wise? Is there anything you can share on how you think about yields of N2 versus N5 and N3 at the same stage of development?
Okay. Rolf's second question is on N2. He would like an update. Are we still on track? What is the timing for N2? Also if there's any update on yields as compared to N3 and N5 at similar stage.
Okay. I said it before, N2 is a very difficult one, but, you know, our progress so far so good. Actually it's a little bit ahead. We are going to have a mass production introduced in 2025. Our customers' engagement so far are very, let me say that comparable with the N3, N5. The status today is very comparable to the N5 and also N3. We are happy to see that, you know, our progress so far.
Does that answer your question, Rolf?
Great. Thank you very much.
Thank you, Rolf. Okay, operator, let's move on to the next participant, please.
Next one we have Charles Shi from Needham & Company.
Good afternoon. Thank you for taking my questions. I have two, both on CapEx. The first question is, really, I understand you are not here to guide 2023 CapEx, but I think a few years ago, you did provide some long-term CapEx, a range of CapEx from when your long-term CAGR guidance was 5%-10%. You think $10-12 billion CapEx is gonna support that long-term CAGR. I think you just reiterated your long-term CAGR to be 15%-20%. Do you kinda have a similar range of CapEx for us to think about over long-term? Again, I'm not asking you about 2023 CapEx here.
Okay. Charles' first question is about really I think, Charles, you're asking about CapEx correlation with growth. He notes that in the past, we, you know, had said our long-term growth would be 5%-10%. During that period, we said we would spend between $10 billion-$12 billion. Without asking about 2023 specifically, Charles wants to know, now we believe we will grow between 15%-20% CAGR in the next few years, what type of CapEx range does that imply or should he assume? Is that correct, Charles?
Yeah. Correct.
Okay, Charles. I think it would be. Let me try to answer this question from the capital intensity point of view. When we invest heavily to capture the future growth, the capital intensity will be high, like last year and this year. If the growth slows down, the capital intensity may become lower. Now, longer term-wise, we think that a normal or a reasonable capital intensity may be somewhere between mid- to high 30%, longer term-wise.
Thank you. The second question is about very near term into the next quarter, basically, because you just slashed your 2022 CapEx, I mean, from what you guided one quarter ago, by about $4 billion. If I hear correctly, half of that can be attributed to the tool delivery issues. The other half, can you clarify a little bit because you kind of said it's about capacity optimization. Is that kinda like a $2 billion reduction in CapEx, mostly a reduction of the equipment spending, or is it something else, you're reducing here and trying to improve the capital or capacity optimization here? Thank you.
Okay. Charles' second question is on the CapEx, and he is asking, we have now guided to around NT 36 billion versus, you know, close to NT 40 billion last time. We have said half of it is related to tool delivery, and the other half is capacity optimization for the midterm demand outlook. His question is, with the capacity optimization, is this mainly also reduction in tools, or is this some other issue? Right. Is that correct, Charles?
Yes. Correct. Thank you.
Okay. Charles, I think you're right, that the half of that difference comes from tool delivery issue. The other half is the capacity optimization. That's because of the current uncertainty in market conditions, so we're tightening up our capital budget. It relates to the whole capacity. It's including tools, including the other stuff within the CapEx. Yeah. It's mainly N7.
Thank you very much.
Yeah. I think.
Thank you. Thank you.
Yeah. It's tools, and also as CC and Wendell mentioned, some of the adjustments we have made to our N7, N6 capacity and CapEx, due to the aforementioned reasons. Thanks. Okay. Thank you, Charles. Operator, can we move on to the next participant?
Next one we have Brad Lin from BofA Securities. Go ahead, please.
Thank you very much for taking my question. First of all, still, congratulations on the strong earnings. My first question is about, well, if we look at the long pages of the new regulations from the U.S. on China, it can be really broad and, of course, with a key aim on the supercomputing. However, given the wide variety and also wider application of chips that TSMC makes, isn't it, in theory, difficult to identify and make sure the chips and the business complying with the regulation, or will there be some extra costs and overhead costs by that which we should notice in the future? Thank you. That's my first question.
Brad's first question is, with the new U.S. regulations, he notes it's very long and very wide in scope. He is wondering if it's very hard to interpret, and therefore, will this result in greater overhead costs for TSMC to ensure that we are and continue to be fully compliant with all the regulations?
Well, actually, although it looks about more than 100 pages, from our initial reading and feedback from our customer, actually the regulation is very simple to be understood, like a control switchboard at a very high-end specification. For example, like a 600 bits per second, you know, the bandwidth or those kind of things. Or it's very, very easy to be understood, and we continue to work with our customer to make sure that we fully comply with the regulation.
Brad would also like to know, Wendell, from a financial standpoint, will we incur more overhead costs as a result?
At this moment, we don't think so.
I see. Thank you very much. Just a slight follow-up. Because the regulation is just aiming at the supercomputing, but I know in a supercomputer, definitely there are some low-end chips used inside. So if we happen to make those and then we are not sure that it is going there, when we produce them, will that cause some confusion or bring some trouble to us? Thank you.
Brad's second question is, you know, in a supercomputer, there may be very specific high-end restrictions, but there's also, you know, companion chips, other chips used. Would that cause us an issue?
The companion chip, yes and no. No restriction or no regulation at all. We rely on our customer to work with their own customer for that kind of a product.
Got it. Thank you very much. Maybe if I may, second question would be, clearly, many countries would like to build their own foundry or build their own domestic supply chain. We have learned that the cost will be the major downside. My question would be, can we charge different pricing that's basically demanded by our clients, right? With higher cost, if they want to buy from U.S. or buy from Europe, can we charge a higher pricing if that's based on the same node? But if we think about the bright side, would you please share how TSMC could well utilize this kind of opportunity to strengthen our competitiveness in the long run? Thank you. That'll be all my questions.
Brad's second question is about, you know, he notes that many countries would like to have domestic semiconductor manufacturing. TSMC, we are, as C.C. said, also increasing and expanding our global manufacturing footprint. With higher cost, will we be able to charge a higher price? I guess that is essentially his question.
Actually, TSMC's pricing is strategic and consistent. All I can say is we will continue to share our value. The values come from the technology manufacturing, and also that our you know, relationship with our customers. Whether it is in a different country or it's in a different place, it's not in our consideration. Again, I would like to say, we are continuing to share our value, and our pricing is strategic.
Okay. Thank you,
Thank you very much.
Thank you, Brad. Operator, can we move on to the next participant, please?
Next one we have Mehdi Hosseini from Susquehanna International Group. Go ahead, please.
Yes, sir. Thanks for taking my question. I also have my first multi-part question. I'm just looking at your customers' inventories that are on a days of inventory are at a 25-year high. It seems, based on your commentary, that the demand forecast, especially looking at the first half, has weakened. It seems to me that the inventory correction is going to sustain to Q2, and most likely, your shipment would be declining sequentially in Q1 and Q2. Just looking at your customers' inventory, is it? Is that a realistic view of the first half? Because you also highlighted the fact that inventory correction is going to sustain throughout the first half. I'm not asking for a guide, I'm just trying to better reconcile your customers' inventory with your commentary. I have a follow-up.
Okay, Mehdi's first question is on inventory. He notes customers' inventory levels are very high. We have talked about our observation from an industry level that demand is softening in, you know, consumer segments. His question is what is the outlook and, you know, do we expect the inventory correction, I guess, to be more notable as we go into first half 2023? Is that correct, Mehdi?
Yes. Particularly, you've always had your wafer shipment up in Q2, but I think 2023 could be an exception and it could probably decline and that's how I frame my question. The question is, could wafer shipment decline in Q2 for the first time in many, many years?
Well, Mehdi, our current forecast, actually, our supply chain inventory peaked in third quarter this year. We observed that the inventory will start to reduce in the fourth quarter, last quarter of this year. We expect you will see its biggest impact.
In the first half, actually, of 2023. The details of the first quarter, second quarter, or something like that, we are not ready to share with you yet, because of, you know, we continue to work with our customer and to understand their demand.
And we-
If I may have just a quick follow-up to that. Does that imply that your customer focus are changing so rapidly that we have to wait for January call to get the final read on the first half?
No, Mehdi, I think we have always in the past we will guide for 2023 and talk about 2023 outlook and first quarter during the January conference, right? I think C.C. has already said.
Oh.
that with the inventory correction, we expect our business to be more resilient during both the down and upturn, given our technology leadership, and that 2023 is a growth year for TSMC. Okay? We will not comment further on
Got it.
First quarter or second quarter .
Okay. My second question, it has to do with depreciation that was down in Q3, down year-over-year and quarter-over-quarter, and this is despite the fact that CapEx was up 65% in 2021. Does that have anything to do with tool optimization? To that extent, as a second part of the question, would you consider converting 7 nm tool to 5 nm and 3 nm?
Mehdi, second question first is depreciation, why, you know, it is down Q on Q and year-on-year. Also consider converting capacity.
Okay. Mehdi, depreciation, we look at it from a whole year point of view. For this year, as I mentioned, we expect it to be up year-on-year by mid-single digits. For next year, it will be meaningfully higher, but we will share with you more in January. Please understand that every year, there are depreciation newly going into the depreciation table, and also there are depreciation coming off of the depreciation table. So what you're looking at is the net result. As to converting N7 capacity to N5, as we mentioned earlier, N7 demand issue is cyclical rather than structural. So at this moment, I don't think we have that kind of plan.
Okay. Thank you, Mehdi. In the interest of time, operator, maybe we can take the last, the questions from the last two participants, please.
Yes, sir. Next one to ask question, Patrick Chen from CLSA. Go ahead, please.
Thank you for taking my question. You talked about 2023 to be still a growth year. Would you say that growth is pretty much the same compared to what you expected a quarter ago, or it is lower? If so, what's driving this lower growth expectation?
Okay. Patrick's first question is about 2023. Is the growth that we see for 2023, how does it compare versus our expectation for three months ago for 2023, and what is driving this?
Yeah. Well, I think it's too early to talk more too much about 2023. We maintain our statement that 2023, we still expect it as a growth year. Okay?
Do you have a second question, Patrick?
Okay. Yeah, that's helpful. Maybe a follow-up. If I may, any leading indicators that you are monitoring that could, you know, help you or help us determine the growth outlook aside from, you know, monitoring the client's wafer orders?
Okay. Patrick's second question is looking with inventory correction. He wants to know if there's any leading indicators that we can look at or he should look at to see when, you know, indication that the cycle is bottoming.
That's really hard to answer your question. Actually, you know, if you look at the whole industry, that's from the smartphone and PCs, and you read all the quarterly report from all major player, you can sense that when it will go up and when it's a downturn. We definitely have some information. Internally, we do the analysis. Today we are taking a very conservative way for our planning. That's all I can share with you.
Thank you. I guess that's what we've been doing as well. But, obviously, we don't have a crystal ball. Well, I don't have any further questions. Thank you for taking my question. Thank you.
Okay. Thank you. Operator, can we take the last participant, please?
Yes. The last one to ask question is Frank Lee from HSBC. Go ahead, please.
All right. Thank you. I wanted to ask, I guess, a question on your N3, N3E. Is the ramp for both nodes gonna be around the same time or is there gonna be some difference in scheduling?
Okay. Frank's first question is on ramp for N3 versus N3E. Is it at the same time?
Yes.
Okay.
Yeah.
Actually, it's not at the same time. Right now we are ramping up by N3, and N3E is supposed to be one year apart, but because of the progress so well, so we might pull in a little bit for two or three months, that's all. They are still not at the same time.
Right. Okay. My second question is, you know, I know you've reiterated your margin, long-term growth margin target being unchanged. I guess given the headwinds you're seeing now, through the first half of next year, you know, with the currency move already gone up quite a bit, utilization rates dropping, is this? Are you gonna see any potential change in the pricing strategy? I know it can't comment specifically, but, you know, you have seen some price increases in the last twelve months or so. From going forward though, would the pricing strategy still be relatively unchanged given the change in the overall market environment?
Okay. Frank's second question is related to pricing. With the inventory correction, and with some of the cost challenges, he wants to know, will there be any changes to our pricing strategy during this correction? Is that correct, Frank?
Yes. Yeah, thanks.
Okay. Frank, let me answer the question. Again, I want to stress that our pricing actually is strategic and consistent. You say that, do we have any plan to change it? No, it will be consistent. All right? You know, not based on any cycle or opportunistic, and we actually most importantly, we always work closely with our customer to provide our value and to help them to win their own in markets.
Right. No, I'm not suggesting there's any optimistic pricing, but just given the market situation has changed for your clients who are probably going through a more difficult time. Would there be some change in strategy to help align with them as well? Or would they largely still be the same policy that we've seen?
We stay the same. We are consistent.
Okay.
Okay?
Okay. Thank you.
Thank you, Frank. All right. 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 one hour from now. The transcript will become available 24 hours from now, both of which will be available through TSMC's website at www.tsmc.com. Thank you for joining us today, and we hope you will join us again next quarter. Goodbye and have a good day.