Advantest Corporation (6857) Earnings Call Transcript & Summary

[Audio Gap] [Interpreted] Thank you very much for taking the time to attend Advantest's 2023 financial briefing. Let me introduce today's speakers, Representative Director, Senior Executive Officer, Group CEO, Mr. Douglas Lefever; Representative Director, Senior Executive Officer and President, Group COO, Mr. Tsukui; Senior Executive Officer, CFO and CSO,, Executive Vice President, Corporate Strategy Group, Mr. Mihashi; Senior Executive Officer, CCRO, Executive Vice President, Sales Group; Mr Nakahara. Serving as the moderator for today, I am Kobayashi, IR Department Manager, Corporate Management Group. Today, first, Mr. Mihashi will present financial results for FY 2023. And then Mr. Lefever will present the FY 2024 outlook, First, shortly, in January this year, Group CEO position was assumed by Mr. Lefever, who will give opening remarks. And then Mr. Mihashi will present 2023 financial results. Over to you, Mr. Lefever.

Okay. Thank you, Kobayashi-san. First of all, I'm very honored to take on the responsibility of CEO. I want to first acknowledge Yoshida-san and the leadership he has provided and the results he's delivered during his tenure. I also want to emphasize the importance of the role of Tsukui-san in the company. In his position as COO, he will oversee many critical operations of the company and provide critical support, both globally and domestically in Japan. Tsukui-san and I have worked many years together in Advantest. I also should mention that Mihashi-san, our CFO and CSO, also grew up together with me in Advantest. Mihashi-san, today, will go through our FY '23 fourth quarter and full year results first. He'll also report on results of our second mid-term management plan, which covered our last fiscal 3 years. Then, in the second part, as Kobayashi-san mentioned, I will cover our outlook for the fiscal 2024 and share some assumptions and market share positions looking forward. With that, I'll hand it over to Mihashi-san.

[Interpreted] Thank you very much. Now let me go through the financial results for FY '23. At the outset, I'd like to give the overview about the results of FY '23. First off, there are uncertain economic environments that continues at this point. I'd like to first appreciate and thank all of our suppliers, employees and all of our stakeholders for their continued support of our business. Now compared to the record results of FY '22, sales, operating income and net income in FY '23 were reduced. But we'd like to highlight that we gained market share in both SoC and the memory testers market. Despite a significant increase in sales for high-performance DRAM related to generative AI, our SoC tester sales declined due to the factors such as softer smartphone market conditions and a slowdown in the server investment. Despite such softer market conditions, the degree of the decline in sales was within the range of our guidance around -- announced in July '22 when we revised the second medium-term management plan, MTP2. However, in terms of profitability, the product mix changed significantly due to lower-than-expected sales of SoC testers for leading-edge processes. FY '23 was the final year of MTP2. We will discuss the results of MTP2 later in today's presentation. Turning to the next page, on Page 5, this is the FY '23 summary of results. Sales, operating income and net income decreased significantly year-over-year. Compared to the full year guidance as of January '24, sales were slightly higher due to the higher-than-expected product deliveries to customers, in particular for memory testers. However, operating income fell short of the full year guidance due to an impairment loss for a portion of goodwill of approximately JPY 9 billion posted in other expenses in the fourth quarter. However, core operating income, excluding one-off items, was JPY 87.1 billion and the core operating income margin was 17.9%. Net income for the year was also below the full year guidance. Tax expenses for the year include approximately JPY 3.8 billion of deferred tax assets in the fourth quarter, which are likely to be realized within a certain period in the future, resulting in a slightly lower effective tax rate of approximately 20%. Combining our year-end dividend forecast of JPY 18 and the interim dividend of JPY 16.25 already paid out, our annual dividend forecast is JPY 34.25. On Page 6, which shows FY '23 sales actuals by segment per region. Over the 3 consecutive years until then, we've delivered many products. And because of the higher demand, we faced a decline in the demand for our products due to excess capacity. And therefore, we had the lower numbers. Now by business segment, although sales for SoC testers for mature process devices such as automotive and industrial was solid, sales for advanced process devices for major customer appliances such as smartphones and PCs for the data -- and as well as for the data center-related applications declined. Memory tester sales exceeded the previous year due to strong tester demand for high-performance DRAM and growth in sales to Chinese memory companies. By region, on a ship-to basis, sales to Taiwan fell sharply, particularly for high-end SoCs. Page 7, which is about the fourth quarter FY '23 summary of results. This slide is describing the results, and the details of the performance will be explained in the subsequent slides. On the next page, Page 8, this is the quarterly sales by segment. Semiconductor and Component Test Systems, the SoC tester sales were JPY 58.7 billion, a decrease of JPY 4.7 billion quarter-on-quarter. And for application processors, sales declined partly due to small pull-forward in the previous quarter. On the other hand, sales for advanced process devices was down only slightly on quarter-on-quarter owing to the sales growth for high-performance computing and AI-related semiconductors. On the other hand, sales for mature process devices, such as automotive and industrial, which was solid in the first half. However, it softened in the second half, resulting in a decline in sales for 2 consecutive quarters. Sales of memory testers were JPY 32.8 billion, an increase of JPY 7.9 billion quarter-on-quarter. In addition to strong demand for high-performance DRAM related to generative AI, sales to Chinese DRAM companies also increased. Mechatronic Systems sales increased due to a healthy number of shipments in SEM metrology products. Service, Support and Others in the System Level Test businesses, which has high sales exposure to specific customers, sales decreased due to a high comparison base following a high level of product deliveries in the previous year. On Page 9, this is the fourth quarter sales by region. As is expected in the previous page, there was the sales pull-forward in the third quarter, as explained. And the sales of SoC testers is decreased significantly on a quarter-on-quarter basis. Moreover, sales of System Level Test business decreased as well. For China, sales increased in SoC testers and memory testers, respectively. In South Korea, robust customer investment continued, and memory tester sales increased. Page 10, this is the fourth quarter sales, gross profit and operating income. Gross margin increased slightly quarter-on-quarter. SG&A increased by JPY 9.5 billion quarter-on-quarter, and this is due to the impairment of a portion of goodwill of approximately JPY 9 billion in other expenses items, as mentioned previously. The impairment is related to Essai, Inc., which we acquired in FY 2019. Due to a lower-than-expected sales outlook of their test socket business to a major customer, the future cash flow projection was deteriorated. The core SG&A was almost flat quarter-on-quarter. On Page 11, FY '23 fourth quarter R&D expenses and others. So this chart shows as is and because this is the end of the year, R&D capital expenditure increased quarter-on-quarter. The fourth quarter FY '23 operating cash flow increased because of the absence of the income tax and bonus payments in the previous quarter as well as a slight decrease in inventories. On Page 12, now as for the balance sheet for the period ended in March 31, 2024, on asset side, inventory decreased slightly compared to the previous quarter. And on the liability side, short-term borrowing was refinanced into a long-term borrowing in view of the working capital. Throughout FY '23, we worked to strategically build up inventories for the medium- to long-term view and to increase supply capacity in the areas where demand is strong and particularly for high-performance DRAM. In addition, inventories increased significantly compared to the previous year, partly in response to push-outs request from customers, especially for SoC testers. To address declining capital efficiency, we will drive our efforts to upgrade supply chain management and seek to monetize inventory in a timely manner. Lastly, I would like to discuss the results of the second mid-term plan, MTP2. Regarding the results of the second mid-term plan, MTP2, as shown in the slide, well, we were able to meet all the original targets set in May 2021. We do not meet the revised target set in July 2022 with the exception of sales. Looking back over the past 3 years, the business environment has changed significantly in a short period of time. In the first half of the period, the increasing digitalization of society drove advancement of semiconductors, while simultaneously pushing for increased semiconductor supply capacity. In the second half, however, the semiconductor market for major consumer applications entered a correction phase. And the tester market, which had experienced rapid expansion previously, ended up slowing down, although there are areas of strength such as high-performance semiconductors related to generative AI. Even under such an environment, we have increased our market share for 3 consecutive years in the MTP2 period by successfully capturing the test demands of leading customers and new entrants working on advanced technology development in the area of high-performance semiconductors of increasing complexity, such as HPC/AI and HBM. Regarding this part, Doug will thoroughly explain it in his presentation part. Despite the short-term cyclicality of semiconductor supply and demand, we see a continuing trend of increasing complexity in the pursuit of higher-performance semiconductors. Including the emergence of new semiconductor market entrants, the semiconductor tester market is likely to continue to grow in the mid-, long term just as MTP2 results exceeded those of MTP1. We recognize that the key challenge for us going forward is how to ensure strong profitability, which can withstand changes in the product mix. With regards to shareholder returns, after taking into account the company's minimum cash-holding level of JPY 100 billion, working capital requirements and growth investment opportunities, we decided to implement shareholder return based solely on dividends for FY 2023. Therefore, the total return ratio was below 50%. However, the total return ratio for the 3-year cumulative period of MTP2 exceeds 50%. That concludes the presentation from my side.

Thank you, Mihashi-san. So let me begin by touching about some of the semiconductor market trends. From our first mid-term plan to our second mid-term management plan, the tester market size in terms of a 3-year average increased by 35%. The market size assumption for our third mid-term plan is currently under formulation. After hitting a bottom in the calendar '23, tester demand will be driven by complexity growth of semiconductors triggered by AI from 2024 onwards, and we expect it to be cyclically up. With the solid high-performance compute and AI demand starting to expand, this year's overall tester market size is expected to be slightly up year-on-year. Due to a slow recovery in end markets such as smartphones and PCs and softness in automotive and industrial equipment markets, it should only be slightly up. In calendar 2025, overall demand for semiconductors is expected to pick up and reach a new record peak, underpinned by a strong demand for advanced logic devices, including those used in high-performance compute, demand recovery for mature SoC nodes and further growth in demand for high-end memory devices. Driven by solid HPC/AI-related demand, semiconductor tester demand is expected to hit consecutive record peaks in calendar 2025 and possibly again in 2026. In both years, further volume growth and performance improvement in advanced logic and high-end memory will also drive demand for our system-level test equipment and test peripherals, particularly for high-end SoC. Now let's look at our sales trend over the past two mid-term management cycles. From the first mid-term plan to the second mid-term plan, our sales in terms of a 3-year average increased by 68%. Our sales growth outpaced the tester market due to market share gains and some partial foreign exchange impact. The overall market expansion is an opportunity for us as we leverage our historical strength in high-end applications such as high-performance compute as well as all-around capability of our solutions portfolio. Our midterm baseline is to secure greater than 50% of the semiconductor test market share, but we will aim for even greater heights, as we have in the past. Now let me turn to our market share position. In calendar 2023, total semiconductor tester market is estimated to have finished around USD 4.4 billion. This is down approximately 15% year-over-year, with the SoC tester market accounting for about USD 3.3 billion and the memory tester market accounting for USD 1.1 billion. While technology advances in high-performance semiconductors drove demand for testers, the market size shrank for the second consecutive year due to the digestion of excess capacity, and so some prolonged effects of the decline in demand for semiconductors and key consumer electronics applications. We estimate our market share was 58%, an increase of about a 1% point year-over-year. We believe we have won the #1 market share for the third year in a row, including a 59% share of the SoC tester market and 56% of the memory tester market. In calendar year 2023, despite the market contraction, we expanded our sales and gained market share in the tester market, especially for high-performance semiconductors, such as HPC/AI and HBM, which are critical markets for our future. Looking forward now into 2024, let me provide some comments on the business environment and test market trend. In 2024, uncertainty remains over the timing of a recovery as demand from key consumer electronics applications such as smartphones and PCs have remained sluggish for a prolonged period, while expectations for an increase in demand remain intact for high-performance semiconductors, including high-performance compute and high-performance DRAM such as HBM. The SoC market tester size in calendar 2024 is estimated to be in the range of USD 2.9 billion to USD 3.2 billion, a downward revision of approximately USD 400 million from our estimate 3 months ago. In the automotive/industrial equipment applications, given the high levels of CapEx spending until recently, we anticipated a slowdown in customer CapEx spending for mature semiconductors and weaker demand for new testers, as existing test capacity gets digested. In calendar 2024, the memory tester market is expected to be in the range of USD 1.4 billion to USD 1.7 billion, an upward revision of about $100 million. In the area of high-performance DRAM, robust customer demand is expected to continue through 2025, in conjunction with increased customer production plans and higher quality assurance leading to high growth rates. Demand for testers for nonvolatile memory is expected to recover gradually in the second half of the year. Taking all these accounts into account -- these factors into account, excuse me, the semiconductor tester market is expected to increase by 5% year-over-year at the midpoint in 2024. Looking beyond 2024 into 2025, we can say with relatively high confidence that the market size will grow. Now moving to our fiscal 2024 forecast. In light of the tester market recovery, our FY '24 forecast calls for an increase in sales and profits with sales of JPY 525 billion, operating income of JPY 90 billion and net income of JPY 67 billion. In the first half of FY '24, sales are expected to decline on a half-over-half basis due to the soft CapEx appetite of customers in the traditional automotive, industrial and consumer applications, despite robust high-performance DRAM demand such as HBM. However, from the second half demand for high-performance compute, such as high-performance compute AI, is likely to absorb excess capacity in the market and drive new tester demand. Gross margin for the full year is expected to decline a bit year-over-year to around 48%, mainly due to changes in our product mix. Operating margin is expected to be 17.1%. With that, we will continue our R&D and investment efforts to enhance the added value of testing by evaluating new test solution offerings aimed at solving customer problems. We are targeting high-growth semiconductor areas that are driven by complexity. This forecast assumes exchange rates of USD 1 to JPY 140 and EUR 1 to JPY 155. Our latest forecast for the impact of exchange rate fluctuations in FY '24 is that operating income is affected positively JPY 700 million per yen of Japanese depreciation versus the U.S. dollar and a negative JPY 300 million per depreciation of yen versus the euro. For our Semiconductor and Component Test outlook, SoC tester sales are expected to be flat year-on-year. Within this segment, for HPC/AI, we are sensing robust demand on the back of increasing production volume and complexity growth for generative AI. Furthermore, demand from new customers is expected. As such, the product mix within SoC testers is likely to turn around. On the other hand, for smartphones, although we expect complexity growth such as 3-nanometer migration to increase test demand, recovery of the demand remains somewhat uncertain as the smartphone conditions remain slow. For traditional industrial and consumer applications, demand is expected to remain somewhat weak this fiscal year. With lead times for SoC testers running now normal -- back to normal to 3 to 4 months, visibility on tester demand is also getting shorter compared for information to us from our customers. This makes it a bit more challenging to formulate the full year forecast accurately. We will continue to work more closely with customers and increase the certainty of our business pipeline. For our memory tester, sales growth is expected driven by high-performance DRAM related to generative AI. However, I should say the current profitability of the memory test business is not at a satisfactory level, and I recognize that improving and increasing its probability is one of the most important issues I need to address as CTO, and I am confident we will do so. Now turning to our Mechatronics, Service, Support and Others segment. In our Mechatronics business, solid sales are expected for our device interfaces and handlers, supported by tester sales growth. Our nanotechnology product sales are expected to increase slightly as well. For sales in Service, Support and Others, they're expected to be similar to the previous year. Within our Service segment, in our traditional maintenance services, demand is likely to be firm due to the steady growth of our installed base. In the System Level Test subsegment, sales are expected to decrease a bit year-over-year. In FY '24, despite the growing customer base adoption of SLT, this business is affected by a certain customer involved in consumer electronics applications, specifically mobility, leading to some soft demand growth in recent times. The System Level Test is still a nascent business area, and I've been working to expand this business, including through M&A. In FY '23, we gained several new customers and new applications such as automotive and compute. The System Level Test business is critical for our strengthening of quality assurance for complex semiconductors, which is an area of high growth, and I remain confident about its future growth prospects. Next, I would like to address key measures for fiscal year 2024. The first key measure for 2024 is to deliver high value-added solutions and expand our supply structure. We will further expand our supply for high-growth devices such as HBM to meet rapidly growing tester demand. We aim to accelerate development of integrated test solutions to enable our leading-edge customers to overcome challenges in developing cutting-edge technology. We will create value through automation initiatives for end-to-end test processes, which coverages ranging from complex semiconductor design, all the way to system level test with the data analytics in between. The second key measure is to drive profitability improvement measures. We will maintain our advantage in our already well-established position in the testing of high-performance semiconductors such as high-performance compute AI and leading-edge memory. Lastly, we will also continue with our work on upgrading our supply chain management to keep up with future demand fluctuation and enhance our production structure. Through the management mid-term plan #1 and 2, which cover the last 6 years for Advantest, we have reinforced foundations for growth, including enhanced product portfolio, expanded business domain and enriched engineering support. We are currently formulating our mid- to long-term management policy and growth paths that are underpinned by the aforementioned achievements. Some of the core strategies are outlined below. There are four strategies I would like to share with you today. Firstly, we aim to continue to outpace the future growth rate of our core market by providing greater efficiency through automation for increasingly complex semiconductor devices. Secondly, we will drive operational excellence to the organization to ensure earnings growth. Third, we will continue to explore adjacent new businesses. And fourth, we will enhance sustainability initiatives with a focus on our net zero transition. We plan to hold a briefing in June to explain our mid-/long-term management policy and provide details, including targeted management metrics. Thank you very much.

Morgan Stanley MUFG, Wadaki-san, please?

[Interpreted] This is Wadaki of Morgan Stanley MUFG Securities. Because a question -- only one question is allowed, and therefore, SoC tester market for the 3 consecutive years is making a negative growth. However, what -- we would like to expect is that perhaps there are replacement needs for the testers. And perhaps [ slab ] is now accommodated with generative AI. And also in the midst of U.S.-China economic conflict, and we have slowed the back-end process capacity, maybe [ post ] to increase in the global market. And therefore, tester investment may be increased. As I think about all these possibilities, I wonder, at some point in the near future, there might be some additions of the sales and performance on your end. What is your take?

[Interpreted] Now in regard to SoCs, Mr. Wadaki, you are fully aware that our major customers is, for example, foundry and OSAT, and such service providers are our customers. And to your question, what is key for us is the improvement of the operating rate and utilization rate. And in FY '23, what we recall, there was an idle capacity and changing the users from smartphone to AI, HPC. And there were so many business for the purpose of upgrading. But in regard to the utilization rate, though it is improving, particularly for the Taiwanese customer, some of the customers are approaching to their full capacity operation. And for the system orders, compared to six months ago, the order level is increasing, as we are aware of. Yet we would like to carefully monitor the situation because there are some foundries who are not on the full capacity yet. Because of this reason, TAM is now being estimated at this point, like you heard in the presentation. But the market condition changed drastically. We need to carefully monitor three months to six months. We need to have a full understanding of the position, and we need to update our outlook. In regard to the smartphones, the sales of smartphones is not as robust as you would hope to be. However, in the second half of this year, our customers will once again migrate to the advanced process nodes. It is very likely that they will advance to the higher nodes. Therefore, TAM is now published in 6-month basis. Probably in the second half of this year, there might be some new trends appearing in the market, as I expect. We will be very careful in monitoring the situation, and we'd like to make a judgment on our own.

[Interpreted] In regard to the global decoupling, like I asked the test [ house ] and OSAT in the United States and the capacity building, what is the impact on that?

[Interpreted] Well, migration of supply chain is actually happening. And because of the decoupling, for instance, from China to Malaysia or from Taiwan to Malaysia that there are some supply chain migration. And the tester demand for us is now becoming visible at this moment. Having said that, though, some of the factories are being built up, and we need to carefully monitor the situation at this moment.

Next from CLSA Securities, Mr. Yoshida, please go ahead.

Hi. This is Yoshida from CLSA. I'd like to ask Douglas about your market share outlook in calendar year 2024 and, if possible, in 2025 as well at both memory and SoC testers. I think the market growth potentials are more favorable for your company, but Teradyne also commented that they are going to take some market share at HBM, but also AI/ASIC. So do you think the recent increased market share can be maintained or further strengthened going forward? That's my question.

Yes. Thank you very much, Yoshida-san. Yes, our outlook for FY -- or calendar '24, I should say, as I mentioned is slightly, slightly up. But as you mentioned, there's going to be a large focus on memory as part of that total tester market for this coming year and into 2025. For 2025, we do expect a higher growth rate than we've seen in the past for memory versus SoC. And into '26 is potential to continue. In June, we will be very -- more clear about the actual market sizes for '25 and '26 and how we're modeling those assumptions. Specifically on HBM, I can comment that we're very comfortable with our position that we have. We've been working with our customers closely for so long. In fact, HBM, I think, we started more than 10 years ago. And so we have very deep relationships, but we also understand that we have to maintain a competitive balance between ourselves and how we're supplying into the system and manage our marginal profits there. Our first objective right now, our first responsibility, is to make sure that we have adequate supply at the performance and quality levels expected by our customers. And by the way, those customers happen to be our SoC customers as well. And so we have an obligation, responsibility, for the entire industry to make sure that we have adequate supply. So as I mentioned in my prepared remarks, we are opening up our supply chains in order to grasp the opportunity of the HBM and the high-end DRAM side. So while I can't talk specifically about customers or the exact market share levels for '25 and '26, I guess, I'll leave the remark that we're very comfortable with our position at this point. Thank you.

Yoshida-san, is that okay?

Well, do you think that market share commission for SoC tester side is also similar to this in HBM memory tester?

I think, historically, what we've seen over the decades is that when there's large waves that happen, if it's PC, mainframe, moving to mobility and now with this era of AI, we've seen a large expansion in the memory business as the computational architectures have changed, and we're seeing that now. And that's why there's such expansion of HBM and high-end DRAM. Normally, after that, it follows with a large ramp-up in SoC. And so what we're expecting to see now is continued growth in the data centers, in the cloud. And then there's going to be some migration into the edge compute. And so you're going to start seeing more applications of AI at the edge, which then will support the volume growth of SoC in consumer and mobility. So our SoC business is still quite strong in our view, given the cycle, because of the HPC. And as that migrates from cloud to edge, we should see that layered on to our business model.

Next, Nakamura-san from Goldman Sachs.

[Interpreted] I'd like to ask about the profitability and the outlook as well as trends. So earlier in the presentation, 48% is the gross margin you spoke of. And compared with the previous year, it will be down slightly. So in the new fiscal year, the sales is going up, and there will be a goodwill impairment. And so if you exclude that, operating profit is flat year-on-year. Is it just because of the product mix situation? And if that's the case, memory tester profitability maybe lower than we anticipate. And earlier, Doug-san talked about the improvement of the profitability. That is one of the key responsibilities. And what sort of scheduled time line do you have? And what sort of level do you think you can raise it to?

[Interpreted] Thank you very much, Nakamura-san, for your question. First, Mihashi would like to respond to this question.

[Interpreted] As you rightly pointed out, FY '24 profit may look slow. The reason why this is the case is because of the product mix. I cannot just say it is all because of product mix. But as for this year, in FY '24, memory tester sales ratio will be very high. That is our projection. Normally, 20% to 30% is the memory, historically speaking. But as for FY '24, this will be much higher than that 20% or 30%. So the product mix will be the largest factor that can impact our performance as well as profitability. And as for profit and as to how we can improve the profitability, first and foremost, in relation to memory, especially HBM, the generation changes speed compared with DDR, commodity DDR. Unlike the generation and creation of business, we have to work with our customers with a very hands-on approach to deliver new devices without depending on [indiscernible]. So the cycle time will be faster, we think. So from the second half of the year, HBM3 will come in, so the generation change will happen. So on that timing alone, with the evolution of devices on our side, we'd like to deliver higher value-added system configuration, and our technologies will be used. And as a result, our profitability should improve. I believe we can expect that to happen. And earlier, Mr. Lefever, Group CEO, mentioned that to improve our profitability, we will strengthen our supply chains to a great deal. And then our procurement and costs will be managed very thoroughly. And also our production will be made more efficient. And by doing so, we'd like to push up our profitability, such activities that will be conducted this year.

[Interpreted] In that sense, HBM sales will go up going forward. And as for FY '25, do you believe that the profitability will go up in FY '25?

[Interpreted] Yes, that is our assumption.

From Jefferies Securities, Nakanomyo-san.

[Interpreted] Nakanomyo-san, one question. Impairment is my question. Is that coming from the large accounts, where their sales forecast was changed, as I understand from your presentation? But was that significant enough to pose a large amount of impairment? And what is your forecast? Do you need to continue to incur such a loss of sales going forward? And also if that particular customer is not the one to sell to, I believe that you still have some potential to expand the business. But yet, you are posting the impairment. Would that be just a conservative approach? Or why did you pose such a large impairment?

Thank you for the question. This is Doug. I'll do my best to answer the question. I think I understand. The impairment that we took, yes, was related because of some overexposure to a certain large customer. And we have done -- we do not have plans to take any further write-downs. We do remain confident in this particular business, which has to do with high-end sockets. And as we transition to a broader customer base from their historical business that we acquired, we should see future business and cash flows related to that division. And so while it's painful as we do that transition and the sales cycles have been somewhat longer than we had originally anticipated, we're encouraged by the opportunities that we have been working on with qualifications and customer expansion. So I think, hopefully, I caught the question right, and that gives an appropriate answer.

The next question is from Macquarie Capital Securities, Damian.

Can you hear me?

Yes, we can hear you. Yes.

All right. This is Damian Thong, Macquarie. I just have a question on the profitability improvement that you're looking for in memory testers. Are you able to say what kind of actions you could take? Because it looks like, given your expectation of strong HBM and memory tester growth next year as well, maybe stronger than SoC testers, it would seem that, actually, there is obviously risk that your margins would still be under pressure next year. But any hints you can have on what kind of actions you can take would be welcome.

Yes. Thank you, Damian. Without getting into really deep-dive details, I would probably put it into the three buckets that Mihashi-san mentioned in his answer, which is that we were looking to achieve better value for our customers by improving their yield and throughput and as well as the technology transitions. So as everybody in the industry knows, the yields of HBM are under lots of pressure. And this is a reason why test capacity is required. But there's also opportunities to work with our customers, to add value to their operations. And if we do that correctly, we should be able to improve our product value and, therefore, our margins. The second thing I would say is that we are working -- co-working, with our supply chain. And so you can imagine the number of suppliers that go into the memory tester bill of materials and the extensive supply chain for productions that we have. So we're working with them closely. And then the last thing is just driving overall efficiencies. So I think, when we look at the detailed plans that we have, we're fairly confident that we're going to have some margin expansion in the memory area. And so it's going to take a lot of work, but we've already seen some of the fruits of the labor from this past year already looking forward into our model.

Next, Okasan Securities, Shimamoto-san.

[Interpreted] Can you hear me?

Yes, sir, we can. Please go ahead.

[Interpreted] So from my side, about SoC and tester, projection has changed so much from three months back. So you have provided an explanation. But once again, which part has changed dramatically compared with three months ago? And if you provide a breakdown between first half and second half, are they much different from each other? For example, you want to be strong in the second half or -- in the second half of '24, seeing as you are going to be too strong yet, what is your projection?

[Interpreted] This is Nakahara of sales speaking. First of all, as I mentioned earlier, in the market, utilization improvement has been slow compared with our projection and expectation, and that is because of the smartphone market. And another point is that now automotive industry equipment machinery tester demand has slowed down somewhat. And there is more clarity about it. So including such developments, TAM projection has been revised downward. And as for first half and the second half breakdown, at some point in the future, our SoC tester utilization will be full in the market. So in the second half of the year, we expect more improvement. Improvement will be clearer in the second half of the year. Did I answer your question?

[Interpreted] Yes. And by the way, AI, out of SoC, is AI in line with your projection? Some companies are saying that AI demand is strong. What about your outlook?

[Interpreted] Generative AI in FY '24 and '25, our tester business, this will be our core. So both memory and logic, both of them will be the key for us. So once we reach a peak utilization, then a system order will come in, in relation to generative AI.

[Interpreted] This is Yamamoto of Mizuho Securities. Can you hear me?

Please go ahead.

Now, in the same point, Mihashi-san, I'd like to ask a question regarding the profitability. And product mix worsening is fully understood. However, would that be the single only reasons of deteriorating operating margin? Share is gaining in steadfast manner. And SoC tester and the memory tester, when you look at the breakdown further, then, as an outcome of a competitive landscape compared to five years ago, are there any products having a lower margin? Is that simply that the product with a lower margin increased? Would that simply the product issue? Or when you look at each line of your products, margin isn't that deteriorating. Could you comment on that, Mihashi-san?

[Interpreted] Thank you very much for your question, Yamamoto-san. To be succinct, for FY '24 -- '23, the profitability was lesser. And business with a lesser profitability have expanded, and that is the main reason. In regard to each and single level of our products, we are quantitatively monitoring the margins, and profitability worsening is not an observed impression yet, basically. Well, for example, wafer tester is mostly in the memory area. And therefore, profitability is deteriorating. But for the other line of products, foreign exchange rates and cost of goods and inflation as well as the wage increase. Of course, we've been receiving a certain pressure of margin deterioration. And yet, we are able to secure a certain profitability.

[Interpreted] I'm sure that you are looking at the Teradyne's number. Their business is slightly different. However, when it comes to tester, their profit deterioration for the tester business is actually smaller than your company's. What is your take?