United Microelectronics Corporation (2303) Earnings Call Transcript & Summary

Welcome, everyone, to UMC's 2023 Second Quarter Earnings Conference Call. [Operator Instructions] For your information, this conference call is now being broadcasted live over the Internet. Webcast replay will be available within an hour after the conference is finished. Please visit our website, www.umc.com, under the Investor Relations' Investors/Events section. Now, I would like to introduce Mr. Michael Lin, Head of Investor Relations at UMC. Mr. Lin, please begin.

Thank you, and welcome to UMC's conference call for the second quarter of 2023. I'm joined by Mr. Jason Wang, the President of UMC and Mr. Chi-Tung Liu, the CFO of UMC. In a moment, we will hear our CFO present the second quarter financial result, followed by our President's key message to address UMC's focus and third quarter 2023 guidance. Once our President and CFO complete their remarks, there will be a Q&A session. UMC's quarterly financial reports are available at our website, www.umc.com, under the Investors/Financial section. During this conference, we may make forward-looking statements based on management's current expectations and beliefs. These forward-looking statements are subject to a number of risks and uncertainties that could cause actual results to differ materially, including the risks that may be beyond the company's control. For a more detailed description of these risks and uncertainties, please refer to our recent and subsequent filings with the SEC and the ROC security authorities. During this conference, you may view our financial presentation material, which is being broadcast live through the Internet. Now, I would like to introduce UMC's CFO, Mr. Chi-Tung Liu, to discuss UMC's second quarter 2023 financial results.

Thank you, Michael. I'd like to go through the 2Q23 investor conference presentation material, which can be downloaded or viewed in real time from our website. Starting on Page 4. The second quarter of 2023 consolidated revenue was TWD 56.3 billion, with a gross margin at around 36%. The net income attributable to the stockholder of the parent was TWD 15.6 billion and earnings per ordinary shares were TWD 1.27. In the second quarter of 2023, the utilization rate was 71%, slightly up from 70% in the previous quarter. On Page 5 is quarterly sequential comparison of income statement. Revenue grew 3.8% quarter-over-quarter to TWD 56.3 billion, mainly due to better product mix lead to a higher ASP. Gross margin rate at 36% is also slightly better than that of quarter 1 and reached TWD 20.25 billion. We have been controlling our operating expenses amid of the current semiconductor downturn. So, the current operating expenses in quarter 2 remained flattish at TWD 5.7 billion compared to TWD 5.78 in the Q1 2023. Operating income reached TWD 15.6 billion or 27.8% gross margin -- 27.8% operating profit margins. Net non-operating income in the second quarter reached TWD 2.8 billion, mainly coming from expenses, interest gains, as well as some investment income from our affiliate companies. After income tax of TWD 2.588 billion, our net income attributable to the shareholder of the parent is around TWD 15.6 billion or 27.8% -- percentage point. EPS was TWD 1.27 in second quarter. On Page 6, is the first 6 months comparison. Revenue declined 18.4% to TWD 110.5 billion, and the gross margin rate was around 35.7% or TWD 39.4 billion. Net income in the first half of 2023 was TWD 32.2 billion, and the net income rate was 29.2%. EPS in the first half of 2023 was TWD 2.58 per share. So on Page 7 is our balance sheet at the end of June 30. Cash on hand is around TWD 163 billion, and total equity for the company is TWD 326.9 billion. On Page 8, as we mentioned, the blended ASP benefit from a better product mix. In a way, it's also a lower utilization rate in 8-inch wafer capacity. So, ASP in the second quarter inch up a couple of percentage points compared to the previous quarter. On Page 9, revenue rebounded for our Asian customers in the second quarter and now reached about 56% of the total pie compared to 50% in the first quarter, and North America declined from 31% in Q1 to 27% in Q2. On Page 10, IDM declined slightly to 21%, and Fabless account for 79% of the total revenue in the second quarter of '23. On Page 11, the revenue breakdown by application didn't change much. Communication remained the same. Computer remained the same. Consumer increased by 2 percentage point to 26%. On Page 12, due to our newly ramped capacity becoming available in our P6 in Tainan fab, our 22/28 nanometer revenue continued to rise. Now it's around 29% of the total revenue. 40-nanometer is about 12% decline by 3 percentage point from the previous quarter. On Page 13 is our capacity -- quarterly capacity breakdown. We will continue to see some mild increase coming out of our 12A P6 expansion in Q3. So the last page of my presentation is on Foundry capital expenditure plan for 2023. For the time being, it still remain at USD 3 billion for the 2023 budget. So, the above is a summary of UMC results for second quarter of 2023. More details are available in the report, which has been posted on our website. I will now turn the call over to President of UMC, Mr. Jason Wang.

Thank you, Chi-Tung. Good evening, everyone. Here, I would like to share UMC's second quarter results. For the second quarter, we reported results in line with guidance with wafer shipments remaining flat from the previous quarter and utilization rate of 71%. Second quarter revenue grew 3.8% quarter-over-quarter, mainly due to improved product mix within our 12-inch portfolio. Revenue from 22/28 nanometer product continued to increase sequentially, representing 29% of second quarter sales, while contribution from specialty technology reached 59%. By segments, we saw short-term demand recovery in the consumer space for Wi-Fi, digital TV, and display driver ICs; which demand for computer-related products also moderately rebounded from the previous quarter. We are pleased to share that we have completed the transaction to acquire remaining shares in USCXM, our 12-inch fab in Xiamen, China. As one of UMC's 4 12-inch fabs in geographically diverse locations, USCXM will continue to provide high-quality fabrication services to customers and increase its contribution to UMC's financial performance as a wholly-owned subsidiary now. Looking into the third quarter, wafer demand outlook is uncertain given prolonged inventory correction in the supply chain. While we saw spot of a limited recovery in the second quarter, overall end market sentiment remains weak and we expect customers to continue stringent inventory management in the near term. Despite a weaker-than-expected environment going into the second half, we believe our 22/28 nanometer business will remain resilient due to our strong position in leading edge specialty technologies such as embedded high voltage. In addition, we are gearing up to offer a necessary silicon interposer technology and capacity to fulfill emerging AI market demand from customers. Now, let's move on to third quarter 2023 guidance. Our wafer shipments will decline by approximately 3% to 4%; ASP in U.S. dollars will increase by 2%; Rising costs will erode gross margin by low single-digit percentage point; capacity utilization rate will be in the mid-60% range; our 2023 cash-based CapEx will be budgeted at USD 3 billion. That conclude my comments. Thank you all for your attention, and now we are ready for questions.

Thank you, President Wang. [Operator Instructions] And our first question is coming from Randy Abrams, Credit Suisse.

Yes. I wanted to ask my first question about the shipment outlook, and wonder if you expect, at this stage, for the decline to persist throughout second half? And then if you could split down by application, how you're seeing the auto industrial and then for the areas that started to pick up Wi-Fi, TV, driver IC and PC, how do you see that pickup sustaining? Or do you see those starting to correct now?

Well, I mean, Randy, on a higher level, given the overall softer end market demand, while the post lockdown, the recovery in China has been slower than expected and weakened macro conditions, where our customers are cautiously managing their inventory, and we expect the situation will likely lingering into Q4. So, mainly it is inventory concerns. So, about inventory, while the end market demand has worsened compared to a quarter ago, across segments in smartphones, PC and server, the inventory will be slowly worked off now. The pace of digestion is slower than we previously expected. Although at auto and industrial segment, the demand of year-over-year projection at this time, we -- our projection shows -- remains unchanged from the previous quarter. However, the inventory level has picked up in Q2 for this 2 segment now. So, therefore, we kind of cautiously observe the demand and supply situation now and to determine when the semi cycle will start to improve or come back.

That's helpful. So, to clarify, it sounds like you said the pace of inventory digestion because of weaker demand in smartphone, PC and server. I just wanted to clarify, that's what you were trying to say, like weaker demand that's keeping the inventory correction longer?

That's right, in the segment of the smartphone, PC and server space, yes.

And for 28-nanometer, I think that node has been a bright spot and the ASP guidance seems to imply that's continuing. Could you update on your plan or your goal, I think, to get utilization back up to 90%, 95%, if that's still tracking? And do you see the P6 continuing to get filled up as it comes in?

Well, first of all, for the P6 expansion, we expect our 28 ramp trajectory is still on track at this point. And we remain confident about the -- 22 and 28 nanometer as it will be a [ long last ] node or driven by many applications across the 5G, automotive and IoT. And we have been focused on 22/28 offering, including many of the leading specialty and larger technologies, along with our manufacturing quality and capacity offering now. So I think the 28 and the 22 right now remains resilient. And it's our projection now that 28-nanometer loading will remain at relatively healthy level and mainly supported by the OLED driver, like I mentioned, for the specialty technology and other larger technology like ISP, Wi-Fi and the SOI processors.

For the expansion on track, I think the plan was 12K end of this year. And then the rest of the 27.5K, would that -- what is the timing? Would that be available through next year or any slowdown to move in that capacity?

No, for the P6, the trajectory is still on track. So there is no slowdown. And they are mainly mutually-committed by both the customer and us. And given the current projection, our forecast shows that 28 loading will remain at a healthy level.

The implication then, if you could blend, you have that capacity with LTA on top, I guess, supporting ASP. How do you net out the utilization, looks like now back to mid-60s. So, with the excess capacity, how do you net pricing environment mature node versus the new capacity coming on, if the trend for you to manage too stable or do you expect a bit more price pressure just with the prolonged inventory correction.

We definitely see the pricing pressures. The -- but the blend is actually more a reflect of our loading between the 12-inch and 8-inch as well. So, if we start off this -- your question with the ASP, then it's our belief when the end demand remains weak, the pricing does not help to stimulate the demand. We are respecting the -- we are respecting the overall foundry market dynamics and the pricing pressure, while the UMC will remain our pricing strategy of maintaining our pricing strategy. But we do work with our customers during the good time in a bad time to uphold their competitiveness and relevance in their respective market to secure their market share. In addition, our value-added technology, I mean in fact the quality and capacity alignment will support our customers. They're competitive as well in their market position. So in the pricing space, we respect the market dynamics and the pressure, but -- and we're closely working with our customers on that. In terms of the blended ASP, they are LTA for the P6 on the 28. And there is non-LTA Protected 28 and there are also an 8-inch outlook. So, if we blend it together right now in Q3, the ASP is projected, it's going to be about low single-digit increase. In the forward-looking, we have to give a quarter at a time because that was subject to a different product mix.

Next question is from Brett Simpson from Arete.

I wanted to just come back to LTAs. And maybe if you can discuss whether customers are generally sticking to these LTAs. And if not, can you maybe just talk about how you're resolving contracts with customers? Are you collecting penalty payments in recent quarters? Could you maybe quantify that? Any thoughts in terms of how you're managing through these LTAs given it's a sizable amount and your utilization is obviously coming down quite significantly from where it was a year ago or so. So, any update there would be very helpful.

Sure. I mean LTA is a mechanism to support both customers and us for a longer-term perspective, because we want to make sure -- the customer want to make sure there is supply resilience, and we want to make sure that our investment is also somewhat protected, which is such a commitment. So LTA is a mechanism of that relationship. During the downturn cycle, we both have to look at this and revisit the situation. And -- but under the framework of, we still have strong confidence in our long-term projections. So, there are some modulation or adjustments to be made during -- given the LTA relationship. So, they are resulting in some of the resolution with current LTA agreement. So that results in some of the financial penalty related matter. But at this point, it's still in a smaller fashion. I think with both know customers and us still have confidence in what have we aligned to do in terms of longer term. So, we expect the market dynamics, and we will work with our customers in trying to dealing with those challenges. But again, the LTA is something that we both respect and seriously commit to it.

Okay. Maybe just switching gears a little bit on the gross margin situation. And the guide -- obviously, gross margin is coming off a little bit from a resilient level in Q2. Where do you see the trough in gross margins for UMC as we go through this downturn? Do you think it will be -- gross margins are troughing in Q3, do you think it might -- the weakness in gross margin might extend into maybe early next year? Any thoughts in terms of depreciation as well because it seems to be creeping up a little bit.

Well, I mean that's a multiple level of the consideration. One is the current market pricing pressure. The other is the recent rising costs. For Q3, the rising cost in electricity hike, raw material, labor will impact our profitability, and we will continue focus on actions to mitigate those rising cost headwinds. We have implemented aggressive cost reduction activity to control the power consumption, productivity, manage the material costs and usage and streamline our process flow and labor management via smart manufacturing measures. We'll also focus on technology improvements such as continuous improvement process, CIP, optimize our fab productivity and quality to help mitigate those headwinds. So, we will try to maintain our structural profitability level at a healthy level. But given the market current dynamic, we're probably giving -- will be more appropriate or prudent to giving this margin guidance 1 quarter at a time. And once we're over the downcycle and then we'll probably can share a bit of a longer time perspective.

Yes. So the depreciation numbers for the 2023, we're still looking for a little bit over 5% year-over-year decline. And for 2024, the increase will be more meaningful. But we -- again, we'll offer the numbers when we approach to the year-end.

Next question, Charlie Chan of Morgan Stanley.

So, my first question is actually about your 14-nanometer strategy or progress because one of your defense service partner just mentioned that there will be 14-nanometer AI ASICs demand and production could be in 2 years. So I'm just wondering that whether company is already prepared for the 14-nanometer demand? And any more details about the capacity expansion, production timing, et cetera, will be much appreciated.

Sure. Our goal is always to fully exploit the DoD capability. And as we know, the DoD capability can sustain to the FinFET technology. And having successfully enter into mass production of our 22-nanometer business now, and while we're witnessing the steady rise in revenue contribution, we are actively progressing with development of the 12 FinFET, and specialty FinFET based on the existing 14-nanometer technology node. So, from the technology development, we are actively progressing. However, on the capacity side, we still need to align with our customer for the capacity expansion and subject to our ROI justification principle and which we have adopted for years now. So, once that's more clear, then we will update accordingly.

Yes. Just to add on to that. We do have 14 capacity already for years. So, we have been producing some 14-nanometer crypto coin-related product a few years back. So, I think what Jason was referring to is more massive capacity expansion going forward. So currently, we already have some 40-nanometer capacity already.

Okay. And my next question is about your gross margin erosion in third quarter. I know that -- mentioned about the cost increase from electricity or raw material. But I'm wondering, maybe this question is to Chi-Tung, do you consider the -- any dollar depreciation, I think -- I would think there's a kind of a tailwind to your gross margin in third quarter. And also, I wanted to know what was the kind of impact from the pricing erosion. So, can you give us some more details about; #1, is the currency depreciation, and secondly, the pricing pressure.

Yes. For our guidance, we didn't factor in the currency fluctuation as a factor. So they won't be included in the couple of percentage erosions in our third quarter margin guidance, which is not included.

Okay.

Yes. As for ASP, our blended ASP guidance is up by 2 percentage point. Of course, like-to-like, it may vary according to different notes. And they also got some kind of mathematical help due to the lower 8-inch wafer capacity utilization, right? So, again, it's already reflected into our gross margin guidance.

Okay. So Chi-Tung, where there is -- for every 1% difference of the TWD depreciation, what does it mean to your gross margin benefit?

It's about 0.4% at this point.

Okay. Yes, and next is more focused on this pricing, so they have a [indiscernible]. So I think your peers more or less, they already mentioned some here pricing pressure instead, I can understand. And I think you should be appreciated by the customers that you want to support them to maintain their market position. But my concern is more about the 28-nanometer, your pricing strategy, because I mentioned that your 28-nanometer utilization is still healthy. You still have those AMOLED ISP demand. So I'm wondering for 28-nanometer, are you also being flexible about pricing because we keep hearing some of your customers talking about 28-nanometer, you may have some compromise as well.

Well, I mean our pricing position are the same to all nodes, although the different node has a different circumstance or situation, but our position remains the same. We're maintaining our pricing strategy, but we do work with our customers, whether on a node-to-node basis or at the holistic level, and -- even in good time or bad time. I mean the end goal is to try to help them and support them to be competitive and secure their market share. So, it -- that's not separated by different technology nodes.

Okay. So it's not just limited to the inch, is that right?

No. I mean is it in general principle, our pricing position and strategy, yes.

Yes. Okay, so in that case, can you give us a sense, how much of your business is taking this kind of a flexible pricing? Do you think more than half of your -- knows your business are being flexible about pricing in second half?

I mean, it's hard to quantify that. And -- but given the weaker end demand, I don't think any pricing changes will actually stimulate the end demand. It's mainly about the -- within the pie, modulation between different players. And we definitely look at that closely and working with our customers closely as well and to support and make sure that we secure the share for both of us. And so it's -- from a percentage or quantitative point of view, it's hard to give it out at this point.

Okay. And lastly, maybe following the Brett's question about the trough of gross margin. I'm wondering your view about the pattern of the fab utilization. So third quarter, you see a sequential decline of UTR, right? Do you think that is the pattern of the utilization for this cycle?

Yes, it's difficult to predict, and we've also did that. So, what we can say is the current weakness, there are inventory digestion, the slow pace, it's going to linger into Q4. I think that's all we can say for now. And we're certainly happy to give the gross margin guidance for the next quarter at the end of July.

Okay. So, definitely, we want to listen to your opinion, but my observation is that your customers, although their demand is slow, but they are outgrowing foundry sectors, right? So I do think the same inventory will come down. And also don't forget that the channel inventory or downstream inventory is a [ present ]. So I'm optimistic. But anyway, we look forward to your next update.

Well, great to hear that.

Next one, Bruce Lu of Goldman Sachs.

I think, to be honest, and we are still surprised about the CapEx remain unchanged and your run-down schedule for the P6, it may not change. I mean, Jason seems to be -- you mentioned like confidence about where your customer is. I think that's the case between investor and management. Can you help us like what gives you the confidence when you have a lingering fourth quarter and your customer cancelled orders all the time, and you -- do you believe that you can maintain those with high retention rate for 28-nanometer of 0.4% to support the CapEx?

I mean, first of all, the 28-nanometer is serving some of the important applications. And for the OLED driver, the current penetration in OLED end market and the volume remains healthy. And in the ISP and the Wi-Fi space, we see new application continue coming into the space? So for the 28, given the broader customer base and diverse product line and product pipeline, that gave us the confidence because the outlook remains healthy. And our -- that also equipped with our specialty technology leadership position as well. So, I think that's differentiating us with other 28-nanometer capacity out there. Right now, given the market outlook, we have been very cautiously look at it because the -- across all segment, the sentiment is weak. But what we have received and what we have validated the -- and what we have delivered right now that has demonstrated the current broader customer base and the product pipeline does give us that confidence that this will probably stay in a healthy level for some time.

So committing something like your 40-nanometers doesn't going to drop a lot, right? Can we compare some of the 40 to 28 or to save some CapEx or what's the schedule for the Singapore new fab? Can we slow down a bit for that, even though it might not impact our CapEx for this year, but we can have a more conservative CapEx outlook for next year?

I mean, we definitely look at this. Right now for the P3, we expect it will go into volume production by mid-2025, as planned. So there's still kind of 2 years out. So we will monitor in the overall market and align with our customer for the P3 ramp. Once we have a further update, we will revise accordingly. And since most of the CapEx increase, now the P6 is pretty much done, and it's more associated with the P3. And I think it's [ less affect ] our 2023 number, but there is a possibility that for 2024 number, we will continue to assess and update.

So, another thing I want to switch gears to, you mentioned that you will do some interposer business. Can you help us understand your strategy for this interposer business? What kind of capital intensity, what kinds of return is going to be, do we need to expand the capacity for that? Is that be a bottleneck to impact your other business? Can we have more color on that?

Sure. For the interposer, as part of our advanced packaging space, we have been providing interposer and wafer-to-wafer asset bonding technology for years. So it's not something new. As the increased demand for higher bandwidth and reduced -- a smaller form factor requirement, we have invested in the space, and we will not be absent on those emerging markets. At the same time, it's our strategy for this space to work with the OSEP partner and to enable an open ecosystem. So sort of, we are only providing the interposers within the supply chain, and we're working with OSEP for the back-end process. And so, that's kind of how we position ourselves within this advanced packaging space. The interposer I mentioned earlier, accelerating is really giving the recent AI coverage. We are just gearing up to offer additional capacity necessary to support the customers the needs. And that capacity is associated with the silicon interposer a lot. And the current capacity size is about 3,000, and it's our goal to double that capacity by mid next year.

Okay. Just to be clear that you have no plan to invest in chip on wafer or other package matter. It's only for the interposer for this, is that correct?

Right. Chip on wafer means we are providing wafer, the W.

Yes. No, you are not going to provide any like bonding or chip -- bonding, de-boding, put on substrate that kind of packaging matter. No, just for the silicon interposer, is that right?

Yes. But we still do wafer to wafer bonding, hybrid bonding, yes.

Yes, our approach, it's an open ecosystem. So, we want to leverage our partners for the downstream and it will be a total solution, but a joined effort by all the ecosystem players.

So what's kind of profitability and return profile for this business?

Rather -- I mean currently, given the size of this capacity and the space, the revenue contribution is still relatively small.

So, how about the profitability?

It's aligned with our current corporate average, yes.

Okay. So you saw it's inline with the current corporate average in third quarter. You said your corporate average went down a lot.

Our corporate average in quarter 2 was 36% gross margin, slightly better than Q1.

Okay. So basically, that interposer business is around that mid-30% gross margin.

But it will not dilute our current corporate average, no.

Next question, Szeho Ng of China Renaissance.

I have 2 questions; #1 regarding the Xiamen fab. Right now is 100% owned by the group. So would there be any strategy change in our China operation or in dealing with local customers there?

I'm sorry. again, the question?

The Xaimen. I mean, right now the Xaimen fab is fully owned by us.

Yes. prior to the buyback, it's actually already dominant or controlled by UMC anyway. And of course, right now, it's 100% owned. So there will be no minority interest once Xiamen continue to make profit. And it's also our goal to see if there's any synergy we can generate between our 2 operations, Hejian in Suzhou and USC in Xiamen to see any more synergy and also competitiveness by these 2 joint efforts. But I think right now, the domestic demand occupy even more percentage of the overall revenues.

See, whether it's a Xiamen facility or our Japan facility, they are 1 of the our 4 12-inch fab under the UMC Group's global capacity scale. With each one of their unique manufacturing location, China, Japan, that will actually position us to support our broader worldwide customer base and with UMC's overall comprehensive technology offering. And at the same time, the UMC's worldwide customer can also access to those local manufacturing site to serve the local supply chain. So we think the geographical locator on those fab actually gave us that benefit as well as our customer to access to the local market. And the 100% acquired or before that, it was the same position that's supporting our deploying customer.

And my second question on the gross margin. So, right now, we are guiding gross margin back to maybe mid-30s level. How easy for us for sort of the company to see the gross margin back to the mid-to-high 40s last year.

I mean, I think we kind of touched a similar question on a few different ways. I mean this current gross margin guidance was given based on the current outlook, our product mix and the loading and the utilization projection and ASP assumptions. And given the rising cost, there are lots of factors that will affect the long-term gross margin projections. One way to look at this is if we look back, I think compared to the UMC in the past, if we look at a similar loading situation as well as the cost structure, I think we are much more resilient now in terms of our financial gross margin level. And going forward, it remains the same. We're going to take the same approach and continue focus on the [ us control ] cost down path, productivity improvement, all the necessary measure against this pressure headwind. And hopefully, we can deliver much healthier, stronger balance sheet for the company. Now, since we're going through this down cycle now and it's more appropriate and prudent that we don't give out any long-term projection at this time. And -- but once it's more ready, and we'll definitely will share that with you.

Alright. Last one. And just to update the industry outlook for the year, the addressable market that we are serving.

Sure. Well, the semi outlook, we expect the 2023 semiconductor market excludes the memory will decline by mid-single digit year-over-year. For the foundry, we now expect the industry will decline by mid-teens year-over-year. With the weaker macro conditions, we'll need to be very conservative as our customer continue to manage their business and inventory. Now, for our addressable market, I think we'll be higher than the mid-teens, yes. The decline will be [ higher ].

Next one, Gokul Hariharan of JPMorgan.

My first question, I just wanted to ask a little bit more on the pricing strategy and what you're seeing from price pressure. Are we starting to see more price pressure coming through for your China facilities given we hear about a lot of foundry price pressure in China, given that UMC has significant capacity in Xiamen and Hejian. Is there a bigger price pressure that you're seeing for your Chinese capacity? Or is it a price pressure that you are seeing across the board for the company itself?

It's across the board. Given the current market condition, and I think we're seeing pricing pressure across the board, not only from China factory, no.

Got it. And we do hear that many of your existing clients are considering using some of the Chinese foundries at least for a portion of their future products. How do you see the China capacity build, because that seems to be the one area where there doesn't seem to be any pause in capacity build-out. Still seems to be pretty aggressive among all the Chinese foundries for primarily mature 12-inch given they cannot really build leading edge. So Jason, how do you expect this to kind of interact with the -- like price discipline that has existed in the non-China part of the market, including you guys and some of your peers over the next couple of years?

Sure. Well, I mean, first of all, I mean, without commenting about the peers, for UMC, to stay competitive and remain relevant in our industry, we have established several advantage in our view for many years. One is the comprehensive specialty technology offering. 2, the competitive world-class manufacturing quality and our geographically diversified manufacturing sites. In addition, our strong commitment, bringing this company to improve our customer relationship and ESG commitment, we believe will further enhance our position as more of a trusted foundry partner. That's more on a higher level. And on the technical level, we look at the major overlap area in the 8-inch. We have been improving our 8-inch customer stickiness by aligning with our customers on their product back, differentiated via specialty technology, including process customization, JDP, the joint development program for products such as analog, power management, IC, MCU and discrete devices. And so, we are confident to navigate through this current market condition as well as the competition landscape. I think there's many areas that we need to do, but we do believe that we have several advantage in many areas. And we definitely have deployed those initiatives, and hopefully, we can navigate through this.

Okay. Maybe one other question. I think you have consistently mentioned over the last few quarters that cutting price or offering price discount in a downturn doesn't really work. Now, as we think about potentially going into next year, emerging from the downturn, do you think that's when we start to see a little bit more price aggression given a lot of your peers as well as you would be running at lower utilization? Is that when we should start to expect a little bit more price accretion in this industry?

I mean, obviously, we all know, right? I mean, the price pressure is there and our comments really about, the pricing doesn't help to stimulate the end demands. So the -- when the end market is shrinking, the overall expense shrinking, the pricing is mainly for the tactical level of approach and which we will use. We would not ignore that. And we -- like I said, we respect the overall foundry market dynamics and those pricing pressures. So we will work with our customer for those to help uphold their competitiveness and secure their market share. And on a pricing strategy level and we remain clear about how to manage those in a separate level of whether it's a strategic level or a technical level or the short-term or longer-term structure. So we -- those strategy remains, no change. But we will definitely use those if it comes to more of a technical purpose. And so we are not ignoring that, we would just want to managing the pricing prudently.

Understood. Maybe one question on 8-inch. It's been kind of seeing very low utilization across the industry. Do you feel that that's going to turn around sometime soon? Or there is a situation in the market where more and more 8-inch product is getting converted into 12-inch. And as a result, this overcapacity situation could last for a fairly long period of time.

I mean for the long run, we foresee some 8-inch demand will recover post inventory correction. They will have some. There are also new applications from megatrends such as EV plus increasing IDN outsourcing business, which will help lift the 8-inch loading. However, like you said, we do anticipate continuous pressure from some of the 12-inch mature nodes that has impacted the 8-inch supply chain. So they will have some recovery and from the overall market once the post inventory correction as well as the new application ramp. But I think what you mentioned about, it also happened there. So that will have some impact as well.

Next on Sunny Lin of UBS.

So my #1 question is on pricing for your long-time agreement, especially for 28-nanometer. So as you said, given the ongoing demand uncertainties, I wonder for your LTAs for P6, and the future Singapore expansion, are you seeing any pressure from clients to renegotiate on the contract pricing? And the second part of that question is for your Singapore fab. If we look at the cost, how much higher is it versus the P6 expansion in Taiwan, it will not be reflecting to your pricing for the contract as well?

So for the LTA, I answered earlier that we -- our customers, as we view that seriously, and at this point, the changing on those LTA is very minimal, relatively small. And between us and the customers, we do look at that as for more of a long-term perspective, not a short-term tactical mechanism. So, we still feel confident about those LTA going forward. They are pricing [ discussed ], but not associated with the LTA. Even with some of the LTA revision, that will be very -- in a very small portion, relatively small. The cost increase in Singapore is definitely much higher, not just because the geographical reason, it is also because the continuous inflationary cost increase. So we -- at this point, we don't foresee that will stop anytime soon. We have to mitigate those headwinds and continue working with our customers to mitigate those, and at the same time, be transparent about the cost increase and to deal with that issue together. At the same time, we have to be realistic about the market price. So it is a balance act, and we'll continue to manage that throughout this whole process.

Got it. And so just to make sure that I interpret it correctly. So, given the cost difference, will it be fair to assume that you could be pressing differently for your 28-nanometer capacity in Taiwan versus Singapore?

For the market price, they will not price differently. The market price reflects the current market situation and it's usually not a cost-based topic. If this is more of a joint investor program, then the cost is more of the factor. But if you're talking about the market price, that's not based on a cost basis.

Got it. That's very helpful. So, my second question is about the structural supply versus demand for growing that foundry. If we look at your historical utilization rate through cycle, that will be between, I think, 85% to 90% through cycle. But apparently, in the last 2 years, the industry capacity had increased quite a bit. And so, just want to get your thoughts on how we should think about the sale cycle utilization rate for you, maybe for next 2 to 3 years.

Well, I mean, from a business management point of view, we also want to help increase our utilization and help load our fabs. So, I mean, from a financial model standpoint and the company's balance sheet healthiness level and the resilience will probably have to plan out different scenarios. But as a business manager, I probably won't tell our teams to shoot for 80% or 90% utilization as our goal. So, from a financial simulation purpose, we have a different layer of utilization rate assumption and to examine our financial resilience. But in terms of the business objective, it's our goal to fully loading our fab.

I see. And so, I guess several analysts have asked the demand question from a different angle. But just want to try one more walk here. And so I understand we are still through the cycle trough, visibility is still not high. But as Charlie pointed out, the industry inventory continued to drop. And so will there be good possibility that we could potentially see a more meaningful recovery from first half of next year?

I mean, I certainly hope so. And I will also expect that if that's the case. So I was glad to hear what Charlie mentioned. And we will closely monitor the market dynamics. And right now, we do see the inventory correction more lingering into Q4. We haven't seen any meaningful demand recovery yet. Hopefully, a quarter over, and we will have a better view and better comments. But at this point, we do think this inventory situation will linger into the Q4.

Next one is from Brad Lin, Bank of America.

So I have 2 questions. First one is pretty short, it's 65-nanometer. So we saw the 65, actually the mix increased from 19% in 1Q to 23% in 2Q. What was the driver behind and should this trend continue in the second half? That's the first question.

I mean, our 12-inch loading is still above our corporate average. And we don't comment about the loading by node. And I can share with you the 65 loading in Q2 is -- increases is mainly coming from the automotive segment.

And then my second question is on the advanced packaging. We would like to learn UMC's strategy and development in advanced packaging. We have learned from Faraday's earnings call yesterday, its collaboration with UMC on advanced packaging on multiple angles. And we also understand interposer and wafer-to-wafer bonding, and the current focus of the firm, how fast do you expect it to grow and do we have any incremental CapEx plans on it? And also, what are the upcoming offers on top of those that can help UMC capture the upside from the advanced packaging?

I do think because -- the higher bandwidth and reduced form factor are the 2 major driver for this advanced packaging needs. But they are for different applications. Usually, the higher bandwidth is more for the AI processors. And the form factor is for some of the integration, wafer-to-wafer integration level to help in the form factor. So there's a different application. So it's important that UMC is not absent from those applications. We do see that market demands increasing, and we've seen there's going to be multiple application coming in. So we want to make sure that we have the technology to serving those markets. And so that's why we have been developing this and providing this interposer and wafer-to-wafer bonding technology for years now. The strategy, also about the ecosystem. We're working -- closely working with OSEP partners for the back-end process, and we are also working with the design services company about the integration, the wafer level design. So they are considered as a part of the ecosystem. Right now, we think this is in the early stage of the production ramp. Technology has been there for years, but the volume production has really just happened with the recent AI momentum. But that's more on the data side, the high bandwidth data side. For the form factor side I think there's other pipeline coming in the next year or so. So we haven't really seen a meaningful or high-volume production yet. The CapEx spending on those space is already planned in our budget. That's not changing our overall CapEx number. It is embedded in that in that number already. So, we think it's not to the level that we need to revise our CapEx to support this.

Got it. Just one very quick follow-up is on the so-called open ecosystem. And then obviously, we are seeing a closed system, which currently dominates the market. So, could you please provide some insights into the optimal collaboration model? Additionally, what may be the preferences of their clients in this regard?

I think the concept is important to provide what you are relevant with. And advanced packaging, in our view, is considered a silicon -- a wafer level integration technology. And that's where we're going to be focused on. For the back end, because the back-end substrates, there will be partners in that ecosystem providing that. The back-end packaging technology also have a partner to provide that service. We don't add much of a value there. So that's why we stick with the wafer level and silicone level integration focus. And so that's how we position ourselves and ensure that we will be in an indispensable position, because the wafer level integration has to be done in a wafer with a fabrication fab. But others, you don't have to. So that's why we had to pick the space that we are more relevant.

Thank you. And ladies and gentlemen, thank you for all your questions. That concludes today's Q&A session. And I'll turn it over to UMC Head of IR for closing remarks. Thank you.

Thank you for attending this conference today. We appreciate your questions. As always, if you have any additional follow-up questions, please feel free to contact UMC at [email protected]. Have a good day.

Thank you. Ladies and gentlemen, that concludes our conference for 2Q23. Thank you for your participation in UMC's conference. There will be a webcast replay within 2 hours. Please visit www.umc.com under the Investors Event section. You may now disconnect. Goodbye.