Advantest Corp
TSE:6857
If you don’t study any companies, you have the same success buying stocks as you do in a poker game if you bet without looking at your cards.
Utilize notes to systematically review your investment decisions. By reflecting on past outcomes, you can discern effective strategies and identify those that underperformed. This continuous feedback loop enables you to adapt and refine your approach, optimizing for future success.
Each note serves as a learning point, offering insights into your decision-making processes. Over time, you'll accumulate a personalized database of knowledge, enhancing your ability to make informed decisions quickly and effectively.
With a comprehensive record of your investment history at your fingertips, you can compare current opportunities against past experiences. This not only bolsters your confidence but also ensures that each decision is grounded in a well-documented rationale.
Do you really want to delete this note?
This action cannot be undone.
| 52 Week Range |
4 214
9 824
|
| Price Target |
|
We'll email you a reminder when the closing price reaches JPY.
Choose the stock you wish to monitor with a price alert.
This alert will be permanently deleted.
Advantest Corp
Earnings Call Analysis
Q2-2025 Analysis
Advantest Corp
In the latest earnings call, Advantest Corporation reported outstanding financial results for the second quarter of fiscal 2024, achieving record highs in sales, operating income, and net income. The operating margin exceeded 30% for the first time since 2022, driven by a surge in demand for advanced semiconductor testers used in AI-related applications. This growth reflects Advantest's effective response to customer needs within the high-performance computing and AI sectors, marking a positive trajectory for the company.
The Semiconductor & Component Test Systems segment saw SoC tester sales reach JPY 109.5 billion, a significant increase of JPY 40.5 billion compared to the previous quarter. Memory tester sales also rose, reflecting strong demand for high-performance DRAM, particularly in high-bandwidth memory (HBM). Despite these positives, sales for mature process products remained weak, highlighting the differentiated demand across Advantest's product lines.
Advantest has revised its full-year forecast for fiscal 2024 upward, now projecting total sales of JPY 640 billion (up from previous estimates), with an operating income of JPY 165 billion and a net income of JPY 122 billion. This reflects continued demand for high-performance compute and AI semiconductors, although growth is moderated by softer recovery in other semiconductor applications. Notably, the company expects a gross profit margin increase to approximately 55%, improved from earlier estimates, due to higher sales and a better product mix.
The company has authorized a substantial share repurchase program of up to JPY 50 billion, representing approximately 9 million shares, aimed at enhancing shareholder returns and improving capital efficiency. Additionally, Advantest plans to increase R&D expenditure slightly to support its strategic growth initiatives, acknowledging the critical importance of maintaining competitiveness in a rapidly evolving semiconductor landscape.
Advantest's management highlighted that while the semiconductor market is solid amid geopolitical concerns, demand for testers in non-AI applications remains sluggish. Looking ahead, the company anticipates a growing need for semiconductor testing solutions tied to advances in AI and HPC. The midpoints of the market sizes for SoC testers and memory testers have been raised to USD 3.65 billion and USD 1.8 billion, respectively, indicating a promising market outlook centered on AI capabilities.
Overall, Advantest's robust second-quarter performance, coupled with strategic initiatives and an optimistic market outlook, positions the company favorably within the semiconductor testing industry. While challenges exist, particularly in non-AI sectors, the strong demand for high-performance testing solutions suggests a resilient growth path. Investors should be attentive to ongoing developments in AI and HPC, which serve as critical drivers for Advantest's future performance.
You don't have any saved screeners yet
[Interpreted] Thank you very much for joining Advantest Corporation's financial briefing for the second quarter of fiscal 2024 despite your busy schedule. I'd like to introduce the attendees on our site today: Mr. Douglas Lefever, Representative Director, Senior Executive Officer and Group CEO; Mr. Tsukui, Representative Director, Senior Executive Officer, President and Group COO; Mr. Mihashi, Senior Executive Officer, CFO, CSO, Executive Vice President of Corporate Strategy Group; Mr. Nakahara, Senior Executive Officer, CCRO and Executive Vice President of Sales Group. Serving as a moderator for today's session, I am Oike of IR Department of Corporate Strategy Group.
Today, following the report of the summary of the second quarter of fiscal 2024, Mihashi will report financial results for the second quarter of fiscal 2024. And then Lefever will present fiscal 2024 outlook before entertaining questions from the audience. We will close today's session at 5:00 Japan time.
[Operator Instructions] Today's presentation materials are posted on TDNet and our website. The audience joining us through the telephone line is kindly requested to download the materials. During today's briefing, we will project the Japanese version of the materials on the screen. The audience who prefers to see English version is kindly requested to download English materials.
Before we begin, I'd like to remind you that today's briefing contains forward-looking statements, all of which are subject to risks and uncertainty that may cause our actual results to be different from those in such forward-looking statements.
Firstly, Mr. Lefever will present a brief summary, including the outline of second quarter fiscal 2024.
Doug, please.
Thank you, Junko-san. Hello to everyone. I'm happy to welcome you all to our call.
Let me start by providing an overview of today's presentation. First, I will give an overview of the financial results for the second quarter of fiscal 2024. Second quarter sales, operating income and net income all posted record highs on a quarterly basis. Quarterly operating margin exceeded 30% for the first time since the second quarter of fiscal year 2022. In response to the robust tester demand for high-performance semiconductors used in AI, we executed on delivering to our customers' test capacity needs, which resulted in record-high sales.
In addition, with the increase in sales, profitability has improved quarter-over-quarter.
Next, let me explain the outlook for fiscal 2024. Based on the second quarter results, we are revising up our full year forecast for FY 2024. I'll review the details of this later in the presentation. But we expect a robust demand for high-performance compute and AI-related semiconductors to continue through the remainder of fiscal year 2024. On the other hand, the recovery in demand for testers used for applications other than high-performance compute remains soft.
Lastly, I will touch upon a share repurchase program. Our Board of Directors has authorized a share repurchase program of up to JPY 50 billion or up to 9 million shares to enhance shareholder returns and improve capital efficiency.
Now Mihashi-san will explain the second quarter results.
[Interpreted] Thank you, Doug. Now this is Mihashi. I'd like to talk about financial results for fiscal 2024 second quarter.
Could you refer to Page 5, please? This shows the results of the second quarter. In the second quarter, our product delivery exceeded our expectations amid the continuation of elevated tester demand for high-performance semiconductors. Along with the increase in sales, the depreciation of the yen against the U.S. dollar also provided a tailwind, resulting in an increase in profit quarter-over-quarter. Details of the results will be explained in the following pages.
Next page, please, Page 6. Now you can see the summary of the results for the second quarter. Let me start with Semiconductor & Component Test Systems. SoC tester sales were JPY 109.5 billion, increase of JPY 40.5 billion quarter-over-quarter. Sales increased due to strong growth in sales of the advanced process products in HPC/AI-related application. In addition to servers, on-device AI applications such as computing and automotive processors contributed to the sales growth. On the other hand, sales of mature process product remained soft. Memory tester sales were JPY 36.0 billion, an increase of JPY 3.9 billion quarter-over-quarter. Demand remained elevated for high-performance DRAM, particularly HBM.
Next one is Mechatronics Systems. Sales of device interface increased quarter-over-quarter in tandem with increased sales of testers.
Services, Support & Others. In addition to steady increase in sales of support service, sales also saw an increase in system-level test business for high-end SoC.
Now you can see on the next page, Page 7. You can see sales by region. For Taiwan, following the first quarter, sales in second quarter increased quarter-over-quarter, mainly due to strengthened quality assurance requirements for HPC/AI-related semiconductors at several U.S. fabless companies. This resulted in an increase in sales to the related foundries and OSATs. The second one is Korea -- South Korea. In addition to increase in sales of memory testers, mainly for DRAM applications, sales of SoC testers also increased for HPC/AI-related application.
Next page, please. This is Page 8. So sales, gross profit and operating income for the second quarter 2024. Gross profit margin also increased quarter-over-quarter due to an increase in tester sales, particularly for high-end SoC applications. SG&A, including all other income and expenses, was JPY 46.5 billion, remained nearly flat quarter-over-quarter.
Page 9. The R&D expenses, CapEx and D&A are illustrated on this slide. R&D expenses remained high following the previous quarter. On the right, you can see cash flow. Due to the improvement in sales and profit, operating cash flow and free cash flow increased.
Page 10. This shows the balance sheet. Inventories are at similar level to those in the previous quarter. While our sales have increased, status of demand recovery varies from application to another. For high-performance semiconductors where tester demand is growing, we are strengthening procurement to increase production. On the other hand, where demand recovery is likely to take some time, we will continue to work on inventory management. This concludes my presentation.
Now I'd like to hand over to Doug, who will go over fiscal 2024 outlook.
Okay. Thank you, Mihashi-san. First, let me explain our business environment. Within the global economy, semiconductors continue to grow in importance. Meanwhile, concerns about the future have continued to rise due to geopolitical risks. The semiconductor market is expected to remain solid, mainly due to demand for high-performance semiconductors related to data center-based AI. On the other hand, demand for semiconductors used in applications other than data center-based AI remains somewhat soft.
Based on this business environment, I will now explain the outlook for the semiconductor tester market in calendar year 2024. As advances in data center-based AI drive performance improvements in semiconductors and growth in production demand, tester demand is expected to remain robust. In response to such robust tester demand, we are continuing to strengthen our tester supply capabilities. As a result, we expect the market size to increase further and have once again raised our market size outlook for both SoC testers and memory testers following our revision in July.
At the bottom, you can see the chart that shows for calendar 2024. In the SoC tester market, we're raising the midpoint from USD 3.35 billion to USD 3.65 billion as a midpoint, and for memory, from USD 1.7 billion to USD 1.8 billion at the midpoint. Meanwhile, the demand for semiconductors using applications other than AI remains soft, as I mentioned. Therefore, the overall recovery in the semiconductor market presents a very -- mixed picture.
Now as we look at the updated FY 2024 forecast, we are revising our full year forecast upward, as shown on this slide. We now forecast sales of JPY 640 billion, operating income of JPY 165 billion and income tax -- income before tax of JPY 162.5 billion and net income, finally, of JPY 122 billion. This upward revision of sales mainly reflects the higher-than-expected sales of SoC testers in our second quarter.
While product delivery exceeded our expectations in the second quarter, we remain somewhat cautious about sales in the second half of the year compared to the first half. This is due to the fact that we have assumed a stronger yen against the U.S. dollar in the second half compared to the first half. In addition, given the continued high demand for SoC testers, we have assumed a sales forecast that is in line with our current product supply capabilities, taking into account our parts and materials procurement. We will continue to work on supply chain management, including the procurement of necessary parts and materials to keep up with this demand.
The gross profit margin for the full year is expected to grow to around 55%, up from the previous forecast of around 52% due to higher sales and a better product mix. Given our strong performance, we will make more strategic investments, including increase in headcount. As part of this, we will also slightly increase our full year R&D expenses.
The exchange rate assumptions for the second half are now JPY 140 to the U.S. dollar and JPY 155 to the euro. The latest estimate for the impact of this exchange rate fluctuation for FY '24 operating income is a positive JPY 1.1 billion per yen of depreciation versus the U.S. dollar and a negative JPY 300 million per Japanese yen depreciation versus the euro. Due to the expected increase in sales from our U.S. dollar-based transaction, we have raised our sensitivity estimate for the U.S. dollar by JPY 200 million from July.
Next, I will explain the details of the sales forecast. Looking at our first segment of Semiconductor & Component Test Systems for FY '24, we expect our SoC tester sales to grow to JPY 324 billion. This is an upward revision of JPY 32 billion from our July forecast. Amidst the increasing complexity of semiconductors for high-performance compute and AI, customer investment appetite remains very robust. Although we have increased the pace of product supply since our second quarter, our product lead time for high-end SoC applications is somewhat lengthening compared to a normalized level. But we will continue to push forward to meet all of our customer delivery requirements as we have been doing.
In FY '24, memory tester sales forecast is now JPY 146 billion, which is an upward revision of JPY 3 billion from our July forecast. Amid strong demand for HBM, we have been progressively working to improve our production capacity this fiscal year. Our memory test supply capability is expected to improve even further in the second half, leading to our expectation for increased sales in the second half compared to the first half.
Now let's turn to our outlook for our Mechatronics and our Services, Support & Other segments. For our Mechatronics segment, we forecast our sales to be JPY 65 billion for FY '24. With tester sales growing, sales of related device interfaces have been raised from the previous forecast. Also, our sales of our nanotechnology products are expected to remain solid.
In the Services, Support & Others category, our sales forecast is expected to be JPY 105 billion. For support services, we expect solid demand due to the steady growth of our installed base. In the SLT business, demand remains somewhat soft due to smartphone business, and also for the test socket business, we see softness from customer exposure in that area.
Overall, we expect a decrease in sales year-over-year. However, there are some very positive signs. With the increased complexity of semiconductors and the rising reliability requirements, there is a trend towards adopting SLT for on-device AI applications such as computing and automotive processors. Over the past several quarters, we have won selections at a number of new customers, ranging from compute to new mobility customers as well as automotive customers.
Next, let me turn to our shareholder return topic. Our shareholder return policy aims for a cumulative total return ratio of 50% or more over the 3 years of the third midterm management plan. Considering factors including our upward revision of our full year forecast, we aim to enhance shareholder returns and capital efficiency through the repurchase program of up to JPY 50 billion. The total number of shares to be repurchased will be up to JPY 9 million over a 4-month period lasting from November to February of next year. Please note that the EPS forecast for the FY 2024 shown on Page 13 in this slide does not include the effect of this share repurchase.
Lastly, I would like to present a few additional updates. First, we have published the integrated annual report for 2024, which was released on October 18. Secondly, we hit a major milestone for our SoC tester. Our 93000 is now celebrating its 25th anniversary from its launch. The highly scalable single platform has long garnered the support of so many customers in our industry. Third, we received the Porter award for the -- 2024 for the first time. This award was presented by Hitotsubashi University Business School, recognizing Japanese companies and business units that have successfully competed through unique strategies. We're very proud of this award. Details of the 3 cases mentioned above can be found on our website. We encourage you to look at that.
Lastly, we hold -- or we will hold an IR technical briefing -- meeting on November 28. This year's focus will be on the SoC test business. Details of the event will be made available at the beginning of next month.
This concludes my presentation. Thank you for your attention.
[Interpreted] Goldman Sachs, Nakamura-san.
[Interpreted] Congratulations on your strong delivery. Regarding your guidance for this fiscal year, you mentioned that high-end SoC tester demand remains robust. That is what you said a few -- that's what you mentioned a few times. But it seems like your second half revenue is going to be down half-on-half. But you mentioned about your supply capability enhancement in procurement parts. And I'm wondering what's the bottleneck you are seeing currently. And in terms of the second half guidance, like what is the upside that we can expect if you are able to solve or address all these bottlenecks?
Yes. Thank you, Nakamura-san, for the question. In terms of the first versus the second half for SoC, we had outstanding couple of first quarters of the year, and they were above and beyond what we had expected. And that had a lot to do with increase in test content needed for AI and HPC devices. So while the numbers right now show somewhat, I would say, nearly flat half over half, the reference point is a terrific first half. And so there could be some upside to that second half as well. But right now, we're very pleased with just continuing the pace, which is an incredible pace of SoC test business for us. So that's one thing.
In terms of the procurement side, we're not seeing any real bottlenecks right now that would be pinch points for our procurement. And so we're managing that. We're managing all of our suppliers. We're working even beyond that to our suppliers' suppliers. So we're -- we've learned a lot of lessons from the previous supply chain crunch through '21 and '22. And so we're very -- we act very paranoid always and making sure that we keep those supply chains robust and having learned a lot of things from the previous go-around. So I don't expect that there's going to be any constraint entirely. It's just something we want to bring up to make sure that we're adequately responding to the demand.
And in terms of upside, yes, I mean what we can say is that we've been continuously surprised by how much additional test content there is. And when I mean test content, it's both test times and additional test insertions. And those things have been -- at times been surprising to us, even to our customers. And so that makes a big challenge when it comes to delivering faster than what we're normally capable of delivering, and we've been able to do that. And that goes to a testimony for our operations and how we manage our inventory in the past several years.
[Interpreted] Next, we'll take the second question -- the next question. Morgan Stanley, MUFG Securities, Wadaki-San.
[Interpreted] Thank you for very strong earnings. In terms of the outlook for SoC tester demand for next fiscal year. So this year is extraordinary. Toward next year, there will be a continued advancement of GPU performance enhancement. And so that could trigger replacement demand, and AI ASIC could also contribute towards next year. And so given these potential drivers, what kind of uplift can you expect to tester TAM next year?
Yes. Thank you, Wadaki-san. It's a very good question. It's true that HP and AI, we see continuing to grow fairly robustly. I would say our visibility is limited really through the first half of 2025. And it's very hard to predict on the TAM growth side because what I mentioned before about the test time and test content, there's some unpredictabilities associated with these very complex AI devices.
That being said, if you look at the architecture of the next-generation devices, the complexity is only increasing. So there's no reason for us to believe that the test content is going to do anything but continue to increase. So when we get to 2025, we're going to be putting a lot of work into our own model and be able to explain a little bit more clarity around the FY 2025 expectations. But it's safe to say first half is looking very positive, and we're cautiously optimistic about the full FY 2025 having some growth.
[Interpreted] Next, we'll take a question from CLSA, Yoshida-san.
Congratulations on very strong results. It is kind of a follow-up of Wadaki-san's question. But could you please give us a color of the tester market outlook in 2025? I know that you said it's difficult to forecast at the time, but if there is any color between SoC and memory tester outlook, that will be really helpful.
Yes. Wish I could give you better numbers right now, but we're still just trying to determine the size and the growth percentages both for SoC and memory. What we can say again is that the complexities are increasing. So we'll see the SoC market continue to expand, we believe, and then the memory side also with the new architectures going from maybe 8 HBM stacks to potentially 12 at 16 high. It's easy to do the math on the HBM and the memory tester side. And so we'll see continued content increase on the memory side as well.
And then, Wadaki-san, I didn't -- he touched on the question, I think, which is relevant for your question also, Yoshida-san, which is about the ASIC providers. And so they're becoming a larger part of the ATE market. And that's really -- the easiest way to think about that is the Magnificent Seven group as it's referred to commonly. And NVIDIA is one of those Magnificent Seven, but they're kind of original chip company. So you could say the magnificent 6 are representing this additional upside. And then the ASIC suppliers that then work with those mag 6 companies are increasing their capabilities and their market share. And so the test TAM, because of those developments, is increasing as well. And we can put numbers to it but probably not until 2025.
Well, just a quick follow-up. Could you please let us know the current lead time of the SoC tester? I think that was shrunk to like a 6-month level, but is that expanding again?
Yes. Thanks. We're at 3 to 6 months. And that's -- 3 is kind of more of the standard, and it just depends on a lot of things. I mean we can -- as I mentioned before, there's test time and test content surprises, so to speak, that come that we have to respond to. And so working through these volatile burst needs for the customers while delivering all the while for consistent purchase orders has been a real trick for our operations team. But I'm happy to say that we're meeting the demand of the customer. So we're not the constraint for the industry right now in terms of the build-out of AI data center. So we expect that to continue, and we're putting a lot of effort internally on making sure that we achieve the test time -- or the test capacity required.
[Interpreted] We'd like to take our next question. Macquarie Capital Securities, Damian Thong, please.
Just like to find out a little bit about the competitive landscape on 2 fronts, both on SoC tester and as well as memory. I think your main competitor, obviously, is investing more in R&D and trying to close the increasingly wide gap between the 2 of your companies. But do you see any sign or any risk that's going to happen next year? So maybe the first way to frame it is that where is the market share going in 2025. And on the memory side, your competitor has talked about how the HBM -- they think the HBM tester sales will decline next year. Would this be true for Advantest? And would that -- if it's different, would that signify also that you're gaining share even in HBM? That's it.
Yes. On the first question around R&D investment, we make adequate R&D. We have never constrained our ability to deliver our products based on R&D. And so it's very -- it's hard to compare us to competitors. We're spending as much as we need to achieve the right technology road maps our customers require. So that's one thing. I wouldn't look too much into that. As a percent of sales, obviously, as our sales expand, the R&D as a percent of sales may drop a little bit. But that's more because of the very good sales.
I would say because of our market share position, too, we're able to leverage that into doing good R&D. And so our hit rate on our R&D is extremely high because we are so connected with -- there are key customers that we develop the right instruments in time for their devices. And so that's a key outcome of our #1 market share position. In our midterm plan 3, we've targeted greater than 58% market share in an expanding market. And we see no reason to adjust that 58% or greater market share target.
On the memory side, we don't see HBM slowing down. In fact, we see some growth in that. And as the industry shifts from HBM3 and 3E to HBM4, we're going to see more DRAM stacks and more DRAM within those stacks. And so long as the data center business continues to grow, then the SoC and the associated HBM is going to continue to grow, and we stand to prosper from that. So I hope that helps, Damian.
Just one small follow-on because clearly, you're growing faster than your competitor in SoC tester. So when you said your 58% market share target, it looks like, at least at the moment, you're tracking well ahead of that. Would you agree perhaps?
I would agree.
[Interpreted] Next question, please. BofA Securities, Hirakawa-san, please.
[Interpreted] This is Hirakawa speaking from BofA. I have a question about system-level test. So in your explanation today, you said that in compute and automotive. Are you beginning to see demand in March? Currently, right now, what is the percentage of these 2 applications that is in SLT business right now? And what's your hope for the next few years or next 3 years?
Yes. Yes. Thank you. I think I understood the question. So on the SLT front, historically, our SLT business has been heavily dependent on mobility, and that still is the case now. So when mobility is slow, inherently, our SLT business is somewhat slow. So we've done a good job of wrapping up a good portion of the mobility market, and we've even collected some new customers in that area.
But what we're also seeing now is opportunities in the compute area. And so as the ASIC providers and as the traditional GPU and AI guys look at ways to further improve their yield, they're turning to SLT, looking at different ways to do things, combining burn-in and SLT steps. And it's all with the purpose of increasing yield because yield is so dominant now for those guys because the ASPs of these devices are so high. They can't afford any yield effects. Either defective parts or even down binning is a very expensive proposition. So more tests, more SLT are being looked at for compute.
In the automotive side, there's more kind of what I would say compute types of applications coming into the automotive. And so everything from your standard infotainment to cockpit sensing to the ADAS types of applications are requiring pretty advanced semiconductors, which then lend themselves to meeting SLT also. So it's slow right now, but we are 100% behind and confident in this business. From a -- let's see, the last part of the question, I think, was -- sorry, help me Hirakawa-san.
Yes. How much -- how big is going to be the contribution from auto and computing in this system-level test in 3 to 5 years?
Yes. So I mean kind of a rule of thumb, and it doesn't always apply, but like 10:1 from an ATE SoC to 1 on the system-level test side. And it's probably -- that ratio is probably even more now, maybe 15:1. But we would expect for every 10, let's say, 93000s, you might see 1 of our SLT systems. And it's dependent upon the amount of parallelism and other thermal complexities and things like that, but just if you're trying to size things, that's maybe a good starting point.
It's very difficult -- one more thing. It's very difficult in that market because there's in-house systems. There's rack and stack systems. And then there's some of the commercial players like ourselves. So it's not always easy to get your arms around the actual market size. But we do see it growing into more applications. So it's easy to be confident that the market is actually going to grow. We don't have any doubt about that.
[Interpreted] Next, Mizuho Securities, Yamamoto-san.
[Interpreted] With regards to just content for HPC and AI, as of July, test content was bigger than expected, which led to your original upward revision. But in terms of the test content that you saw in July or you expected in July, it means that there is an additional uplift in terms of your expectation for test content growth. But in terms of your expectation 3 months ago, what is the uplift that you are seeing currently? Do you have an estimate?
I don't know if you guys got it. Maybe it's better for Nakahara or Mihashi. You guys can address, and then I can add some color.
[Interpreted] So this is Nakahara speaking. So as of July, customers' test time as well as test insertions were already growing in relation to HPC and AI logic. Since then, I would say that in terms of customers' devices, test insertions and tech contents, I wouldn't say it experienced a dramatic change. I would say maybe compared to July, we haven't seen a dramatic change in test content.
I would say that in terms of our discussions with customers, since July, we are working to accelerate to our procurement to meet customer demand and expand our production. And for HPC/AI demand, we are -- we have expanded our supply capabilities in a very short period of time. And we think that our second quarter demonstrates our efforts, which is why our second quarter number was very strong.
[Interpreted] Well, in that case, it means that tester per processor, the assumption is the same. So I guess what happened is that delivery time has been front-loaded. Is that correct?
[Interpreted] Well, so I would say in terms of comparison with our July outlook, the main reason or driver is our supply capability enhancement. But as you know, customers' device road map cycle is getting shorter, and they need to develop new devices quickly. And so from now towards next year, it's hard for us to predict how customers' test time will evolve and change.
[Interpreted] Next, Citigroup, Shibano-san, please.
Let me ask one question on your second half SoC tester revenue assumption. Compared to the first half result, which was JPY 178 billion for SoC test overall, you're expecting JPY 145 billion for the second half. How would you like to break down the second half assumption into Q3 and Q4? Is it like third quarter weighted or fourth quarter weighted? And why?
[Interpreted] Let me respond in Japanese. This is Nakahara from sales. In terms of Q3 and Q4, our logic customers demand changes very rapidly. That's what we are seeing right now. And tester delivery requirement is changing every week. And that has to do with customers' production plans, but that also has to do with overall supply chains, not just us, but in terms of there are different factors like assembly capacity. And so changes are very rapid nowadays. So in terms of the split between Q3 and Q4, it's hard for us to foresee right now. But we do not expect such a huge change. We don't think there's going to be a big step change between Q3 and Q4.
[Interpreted] Well, I think your second quarter SoC figure was about JPY 100 billion. From July to September, like in terms of monthly trend, was it constant? Or was there a spike in a certain month?
[Interpreted] I wouldn't say there was a huge spike in September. I would say that every month, there was a constant effort on our part to ship testers.
[Interpreted] So it's not that like you've pulled forwarded or front-loaded the tester shipment kind of. You didn't overstretch yourselves, right?
[Interpreted] Correct.
[Interpreted] Next, Nomura Securities, Yoshioka-san, please.
[Interpreted] This is Yoshioka from Nomura. My question has to do with profitability for HBM and memory testers. So previously, I think the company said that in terms of upside for profitability, HBM/memory testers margin improvement is one driver. Here with regards to this, in comparison with 3 months ago, do you have higher progress? Or do you have a higher level of confidence toward -- for that? And how much factored in the margin improvement potential in your guidance? So please elaborate on your thoughts around memory tester profitability.
[Interpreted] Thank you. Mihashi will take that question.
[Interpreted] Well, with regards to memory tester profitability, as HBM continues to evolve, our stance is to work on upgrades for existing testers. That's one lever to enhance profitability. There are other ways to do that. In terms of the pace of progress, I would say it's slower than originally expected, is somewhat slow. One reason is -- has to do with adoption, but also the trend for adoption is mixed depending on the customer. So it has to do with the customers' varying trends.
[Interpreted] Let me follow up on that. So your American competitor has talked about the post-stack test. They said that they started delivering after stack test. Is the competitive situation impacting the profitability improvement progress?
[Interpreted] Well, in our business, we always face competition. But with regards to our HBM tester business, we've been focusing on this business for several years already. And so we have been a step ahead of our peer, and that has benefited us -- and that is benefiting us. And we already have a significant installed base now. And so based on that installed base, we can then add value or enjoy configuration upgrades. And we are engaging with customers about technological development. And so we would like to continue to enjoy and maintain our market position. Of course, our competitor, they are aggressive in trying to approach and target our customers. However, we would like to maintain our current strong position.
[Interpreted] Next, Okasan Securities, Shimamoto-san.
[Interpreted] This is Shimamoto speaking. My question is about China. So looking at ASML's guidance, China's deceleration seems to be a clear trend. But in the tester market in China, have you seen any changes? And in terms of sales, the testers you sell for the Chinese market is the main Chinese market smartphone? Or do you also serve HPC market in China?
[Interpreted] I'll take that question. Nakahara speaking. For China, as you can see this in the slide, in comparison to Q4 last fiscal year, the percentage is smaller. But then that's because Taiwan's percentage has grown due to HPC. Look, we are not 100% sure about other SPE supplier situation. But in the tester business, I would say that for the Chinese market, we have not really sensed significant changes.
Well, so I would say our China tester business is very well diversified. We -- for logic, we serve mobile but also HPC/AI and also automotive. And in addition to that, you probably -- as you know, memory customers, memory fabs are expanding. And so we also serve the memory market as well. So I would say that our China business is very well balanced. So I would say that near term, we are not seeing changes.
[Interpreted] SMBC Nikko, Hanaya-san, please.
[Interpreted] Well, let me ask a question on the SoC tester business. So you are strong in HPC/AI. You have been able to supply to meet demand that led to the upward revision. But on the other hand, for other applications, you mentioned that recovery has been weak and delayed. So in terms of the percentage split, automotive percentage is down. But in terms of the mobility, smartphone, what is the expectation for the timing of recovery for mobility/smartphones? If you -- let's say if smartphone demand comes back, you will be able to supply? And so what's your outlook for smartphone recovery? And what's your ability to supply testers for smartphone ability if market -- if recovery takes place?
Yes. Thank you, Hanaya-san. Yes, we're -- like probably the rest of the world, we're very hopeful that the smartphone market returns to growth in 2025. We did see in this year some signs of life, especially for AI-based edge applications like handsets and PCs. And that gave us some good business. But in general, it hasn't been enough, as we know, to drive a replacement cycle.
And so many of the prognosticators are expecting 2025 to be the year where there is a replacement cycle. And it has a lot to do with many big companies and driving some killer applications and things like that. So we're hopeful. And if that does happen, I'm happy to say we do have the capacity available for us to deliver not just at the HPC level but to layer on to that the mobility and the edge-based applications as well.
[Interpreted] Well, I think we are close to the -- we still have time, so we'll take the second round of questions. So right now, we have our hand raised from Nakamura-san, Goldman Sachs. Do you have your second -- could you have the second question?
[Interpreted] Well, sorry, I don't think I had my hand raised, but I do have a follow-up question. So sorry to keep asking you about SoC tester second half revenue outlook. So you mentioned in your quarterly fashion, second quarter was JPY 100 billion and third quarter and fourth quarter is maybe JPY 70 billion. But right now, at this current moment, like are you seeing any deceleration or signs? Like have you like pulled forward or frontloaded delivery too much in the first half that could cause deceleration in the second half?
Yes. I think what we've described before is that we saw a massive step function in test content that happened in the first quarter. And so that -- while some of that is going to continue, that was such a large inflection point that some of that capacity is being put in place. And then as new devices come out, we can't yet determine whether or not the test times are going to increase beyond that. We expect, Nakamura-san, that they still will.
When you look at the AI devices you see in the future, reticle limits being hit, I/O chips being split out into chiplets, you see all the additional HBM stacks. There's new bus architectures. All of that is happening. And there's more kind of a mission mode testing, which, in a way, limits the amount of DFT that's available to reduce test times. So you have all of those kind of complexity factors. And then with the need to have high yields because of the ASPs, there's more willingness to do more tests to avoid having to throw away these really expensive packages. So those are things that we didn't have just a couple of years ago and even last year.
And so now we have this new wave of complexity driving test content. And it's not just test time. Like I mentioned before, there's additional insertions that are being added to guarantee quality and reliability. So putting that all into a model is very difficult, is the problem because it's simply too hard to predict the expected test content. And now test content is far outweighing anything to do with unit volume, and it's harder to predict. So we're trying and we're working closely with our customers, obviously. And as I mentioned, as we get into 2025, we should be able to sharpen our pencils and give a better outlook.
[Interpreted] Well, we are close to the end of the hour, so maybe we'll take just one last question. Macquarie Securities, Damian Thong.
I didn't put up my hand. But anyway, since this is the last question, so just one thing on your policy for share buybacks then. Obviously, I think the JPY 50 billion for this year is quite substantial. If you move into next year, assuming that -- I presume you're forecasting a growth year. What would be a framework for the buybacks going forward? Would you sustain this pace? Basically, would there be an intent to send a signal of a programmatic buyback program?
Yes, we -- what we do is we look at a couple of different things, Damian. And the policy of shareholder returns of hitting a 50% ratio, that is the corporate policy set forth for a midterm plan. So the total average for the 3 years will be 50%. And we'll be tracking above that after this buyback. And then the other thing we put, and this is not a hard policy, but we have a minimum cash policy that we need to maintain for working capital and disruptions that are unforeseen.
So those are kind of the 2 boundary conditions. And then beyond that, we're looking at the growth investments. So we'll make whatever investments are required to support our core business and then to look into adjacent opportunities. And so we do that consistently. And even when we may find those, we always have access to the credit markets to borrow if we need to. So we feel very good about our balance sheet, and it's necessary for us to return some shareholder value in the form of this dividend and then the buyback.
Given current market conditions, is this some type of area that you feel that you need to make additional investments into next year? And this is the last question.
Yes. I think there's going to be things that we do mostly to address the device complexity and the test solutions associated with those things. So if we have a technology gap or we have a partner that requires some investment, we're going to be making those to make sure that we have the test solutions necessary for those new challenges. That's one thing. And then if there's ever any kind of supply issue that requires us to make an investment, then we're going to act on that as well. So we keep a pretty robust list and are constantly looking at that as a management team, and we have some of those things that we're looking at now. That's about as much as I can say on that.
[Interpreted] Thank you very much. Thank you for joining Advantest FY '24 Second Quarter Earnings Call despite your busy schedule.
[Portions of this transcript that are marked [Interpreted] were spoken by an interpreter present on the live call.]