© 2024 Alpha Spread Limited. All Rights Reserved.
www.alphaspread.com is not operated by a broker, a dealer, or a registered investment adviser. Under no circumstances does any information posted on www.alphaspread.com represent a recommendation to buy or sell a security. The information on this site, and in its related newsletters, is not intended to be, nor does it constitute, investment advice or recommendations. The information on this site is in no way guaranteed for completeness, accuracy or in any other way. In no event shall Alpha Spread Limited be liable to any member, guest or third party for any damages of any kind arising out of the use of any content or other material published or available on www.alphaspread.com, or relating to the use of, or inability to use, www.alphaspread.com or any content, including, without limitation, any investment losses, lost profits, lost opportunity, special, incidental, indirect, consequential or punitive damages. Past performance is a poor indicator of future performance.
You don't have any saved screeners yet
Intrinsic Value is all-important and is the only logical way to evaluate the relative attractiveness of investments and businesses.
Thank you indeed for your precious time despite your busy schedule. If I may now, I would like to start Advantest Financial Briefing for the Second Quarter FY 2019. Do allow me first to introduce the participants from Advantest. Starting from the screen side and at the center: First, Yoshiaki Yoshida, Representative Director, President and CEO; next, Atsushi Fujita, Director, Managing Executive Officer; next, Kimiya Sakamoto, Executive Vice President of Sales Group; and allow me to introduce myself. My name is Yasuo Mihashi, Corporate Relations Group, serving as the emcee for the session.
First, Mr. Fujita will explain our financial results for the second quarter FY 2019. Then Mr. Yoshida, President, will explain FY '19 outlook. Then, we are quite happy to receive your questions. We plan to end the session around 5:00 p.m.
Without further ado, Mr. Fujita, please.
Hello, everybody. This is Fujita. I am delighted to make a report concerning FY 2019 second quarter results summary.
Looking back at the industry environment 3 months ago, there were considerable uncertainties in response to macroeconomic weaknesses and the proliferation of the protectionist trade policies. We were cautious about the business environment in the second quarter. In addition, due to the strong first quarter results, we expected quarter-on-quarter slowdown in the second quarter. As expected, reflecting the sluggish demand for end products, many customers continued to have wait-and-see attitude in regard to the tester investment. However, once the impending full-scale commercial rollout of 5G, aggressive tester investment continued among 5G-related SoC semiconductor manufacturers. As a result, orders declined Q-o-Q, but not to the extent that we had expected, coming in at JPY 58.3 billion.
Sales and profit also progressed better than we had expected, and both exceeded the figures posted in the first quarter. Year-on-year, memory tester orders fell due to having the demand in the sector, which was in the bottom phase in 1 year ago. But the decline in sales and profit was compensated for by the strong performance of SoC testers holding results flat Y-o-Y. Regarding operating income in the same period of the previous year, we posted a onetime profit from the sale of business sites of approximately JPY 1 billion. Excluding that, profit for FY 2019 second quarter rose Y-o-Y. In the second quarter, the U.S. dollar moved in a somewhat negative direction against our ForEx sensitivity.
Next on the orders by segment. Semiconductor and component of test systems, JPY 43.1 billion, down 13.6% quarter-on-quarter. SoC tester orders were JPY 33.3 billion. As in the first quarter, demand was mainly for new smartphone SoC devices utilizing advanced processes. The first quarter saw the demand for testers for engineering applications for devices used in 5G smartphones, such as application processors and baseband processors. Going one step further, the second quarter saw investments in testers for initial ramp-up of volume production of these devices. Orders for SoC testers in other areas were generally weak. Normally, tester orders are increasing from the first quarter to reach JPY 9.8 billion. This was due to the continued demand for testers for high-speed DRAM and expanding demand from the local memory companies in China. Mechatronics Systems, JPY 6.8 billion, down 5.8% Q-o-Q. Services and Others, JPY 8.4 billion, down 4.6% Q-o-Q. Annual maintenance and contract orders decrease in every summer. But this year, that was an offset by an increase in system-level test related orders, so the rate of decrease in orders overall was lower than usual.
This page shows sales by segment. Sales of semiconductor and component test systems were up 2.1% quarter-on-quarter to JPY 52 billion. SoC tester sales were JPY 41.8 billion. As with orders, these were mainly driven by 5G smartphone-related demand. Memory tester sales were JPY 10.2 billion. As with orders, these were mainly driven by high-speed DRAM test demand. Mechatronics Systems sales were up 40.6% quarter-on-quarter to JPY 9.4 billion. Device interface sales increased owing to the growth of memory tester sales. Services, Support and Others sales grew 18.9% quarter-on-quarter to JPY 10.3 billion as sales of system-level testers for both SSDs and SOCs increased.
This slide shows orders by region. Taiwan is indicated by orange, and orders decreased from the previous quarter due to order concentration in Q1. This is for testers of engineering purpose for 5G smartphone SOCs.
China is in yellow. And orders increased as the launch and expansion of China's local semiconductor supply chain progressed across the SOCs and memory sectors.
As for sales by region, South Korea, Taiwan and China saw strong sales of testers used for high-end SoCs utilizing advanced processes as well as image sensors, part management ICs, et cetera. In Korea, although memory tester sales remained weak in Q2, overall sales were steady, thanks to the contribution of SoC testers.
This page is about operating income. Gross margin was 57.3% to maintain above 55% despite slight worsening product mix. SG&A was JPY 23.3 billion, declined slightly from Q1. Operating income was JPY 17.7 billion. Operating margin was 24.8%, up 1.9 points quarter-on-quarter. As we are currently adding personnel for future business expansion, it is difficult to improve profitability for the time being. Still, we could achieve operating margin of nearly 25%.
This page is about investment and cash flow. R&D expenses were JPY 9.9 billion in this quarter under review. Depreciation and amortization were JPY 2.1 billion to recognize an increase over the previous year. This is due to the new lease accounting standard of IFRS, which was applied from this fiscal year and a temporary increase of the amortization of intangible assets, caused by our acquisition of Astronics' system-level test business in February. As for cash flow, we recorded JPY 25 billion for free cash flow in this quarter on the review. Due to the balance between orders and sales, collection of trade receivables improved to increase operating cash flow in Q2.
This slide is about financial position and shows the balance sheet. As of the end of September 2019, total assets was JPY 323.6 billion. Cash and cash equivalents increased JPY 23.4 billion from the end of the previous quarter to JPY 134.8 billion. Equity attributable to owners of the parent was JPY 213.6 billion. And ratio of equity attributable to owners of the parent was 66.0%, up 1.8 points from the end of the previous quarter.
That is all for the explanation of the financial results of Q2. Thank you.
Hello. This is Yoshida. I will cover our outlook for FY '19. In terms of market outlook for calendar year 2019, may I remind you that in here, we are looking at the calendar year, not the fiscal year, because the [ data ] for the testers and the market size is available on the calendar basis. I hope you understand this.
First of all, a brief discussion on our business environment. The trade friction between the U.S. and China has not yet been resolved, and inventory adjustments continue in the supply chain of many products. With the exception of individual strategic investment, many customers are not in any hurry to invest in testers. However, test demand for 5G-related conductors has been booming since spring. And based on this, we have raised our calendar year '19 outlook for the SOC tester market to approximately $2.6 billion. Back in July, we said it will be around $2 billion, but we decided to raise it to $2.6 billion.
The memory tester market has been in a slump, but investment from several customers has been up a little. Taking this into account, we estimate that the market in calendar year '19 will be -- was around $550 million. Back in July, we said to the effect that it may come down below $550 million. Since Advantest broad customer base is the principal source of our strength, the current market situation and the customers' investment stance are slightly disadvantageous for us. It is kind of difficult for me to fully explain this, but an overall decline definitely is going to have impact on our market share in certain cases. So if we lose in certain customers, then that is going to have a great impact upon us.
However, if our market estimates are accurate, we will be able to maintain our share of approximately 50% or above in calendar year '19 in both SoC and memory, following calendar year 2018.
In recent years, the driving force behind the tester demand enhancement, test reinforcement and extended reliability and assurance need for the higher performance semiconductors. In fact, volume production of high-end semiconductors is now expanding in the SOC and even in memory, although it is not really noticeable in the [ -- a lot of ] sector yet. Thus, we do believe that there is going to be more room for tester market to further grow from next year onward.
FY '19 forecast. Although uncertainty in the global economy and the sense of a deceleration and end product demand and semiconductor demand continued as before, increased demand for 5G-related SoC testers and conductors has made up for it, boosting Advantest's results in the first half to a higher level than originally than expected.
Based upon this, we decided to revise our full year forecast. We have increased our estimate for sales by JPY 17 billion; for operating income, by JPY 15 billion; and for net income, by JPY 10.5 billion. Because of the acceleration of 5G-related demand, tester investment was very vigorous in the first half in FY '19. We therefore expect that 5G-related orders may slow down temporarily in the second half of FY '19. On the other hand, memory and storage-related demand, which has been soft, is increasing, if only slightly. We are now making comprehensive orders and sales and plans around that assumption.
Gross profit margin is expected to improve by a few percentage points year-on-year from 54.5% to around 56%, up to 57%. Exchange rate assumptions for the second half of the fiscal year have been revised to JPY 105 to the U.S. dollar and JPY 120 to the euro, taking into account the most recent actual figures. For the full year, we assume JPY 107 to the U.S. dollar and JPY 121 to the euro.
Now outlook by segment. SoC and tester business and outlook for the current fiscal year. The major new trend for the progress of semiconductor performance to underpin SoC tester demand is continuing. In particular, our sales in this year have been heavily driven by testers for application and baseband processors and for future 5G handsets.
In the first half, we do receive unexpectedly high volumes of tester orders for these device types from several major semiconductor manufacturers. These orders will convert to sales in the second half of FY '19, so that and sales in our SoC tester business is expected to flat from the previous fiscal year. On the other hand, considering the timing of 5G smartphone production next year, we suppose that 5G-related orders may slow down temporarily in the second half of the year. However, we think this adjustment is going to be just temporary.
The adoption of 5G handsets has still to become widespread, and the performance of servers and game consoles will also be improved next year. Additionally, full-scale demand for the 5G using the millimeter wave band is expected a few years from now. We expect that these major electronics industry trends will continue to stimulate demand for associated testing through advances in SoC semiconductor performance, growth in production volumes, and in further adoption of advanced processes.
Next is about memory tester business outlook. DRAM has well-defined technological drivers for memory tester sales that are less susceptible to memory supply-demand fluctuations such as the shift to DDR5 and expansion of demand for high-speed products for high-performance computer. Amid poor overall market conditions, we have not changed our expectation that tester investment for these advanced devices will start in Q4 in this sector. We originally expected it to start in Q3 at the beginning of this fiscal year. But now, we expect the demand recovery will begin in Q4. Also, as indicated in Q2 orders, we are receiving more customer inquiries for DRAM testers than estimate made 3 months ago. So we revised our full year sales forecast for memory testers upward slightly from JPY 30 billion to JPY 32 billion.
In addition to increasing demand for high-end DRAM, number of layers stacked on NAND flash continues to increase. If bit demand returns, the tester market will see an overall structural recovery. Also in China, the DRAM tester market is beginning to ramp up in addition to NAND flash. We'll watch future development with anticipation.
Next page is about Mechatronics Systems and Services, Support and Others. As for Mechatronics Systems, although we have not changed our forecast of decreased sales in net related business due to the prolonged slowdown in memory customers tester investments, we have slightly raised our segment sales forecast by JPY 1 billion to JPY 33 billion due to upward revision of the forecast for memory testers. For Services, Support and Others, we have increased our forecast by JPY 2 billion to JPY 40 billion. In this area, we can expect solid sales in field services, and SSD system-level test is growing more strongly than previous forecast. In addition, the system-level test business of Astronics, which we acquired in February, is making progress as expected. The revised forecast reflects these factors.
For measures for medium- to long-term growth, let me explain a little about grand design announced in April 2018. Advantest is promoting management in line with this 10-year mid- to long-term management policy. Key measures for the second half of FY '19 remain the same, as explained at the beginning of the fiscal year. In order to maintain and expand our market share from CY '20 onwards, we'll prioritize strengthening engagement with industry-leading key customers and maintenance and expansion of our market share. Customer support request for next-generation 5G devices and future new device development continue to increase. We intend to accelerate our investment in staff, such as system engineers and application engineers and in R&D.
At the beginning of this fiscal year, we announced to increase investment to increase our headcount by about 200 in this fiscal year. But as we see significantly strong demand for support from customers, we will hire more employees than this plan. As for capital investment, we do not consider large-scale investments to increase production capacity as this is not necessary for the time being. Still, we intend to actively implement facility renovation for higher efficiency in production and development and invest in IT systems.
Putting this all together, the investment in FY '19 is expected to be JPY 39 billion. Capital investment will be JPY 10 billion, unchanged from our initial plan. Depreciation and amortization will be JPY 9.5 billion. SG&A expenses are expected to be around EUR 95 million to JPY 96 billion, including R&D expenses.
This is the last page for summary.
In the first half of this fiscal year, performance results exceeded our initial estimate, owing to the increased SoC tester investment. As a result, we have revised both sales and profit forecast upwards. Although we foresee the possibility of a temporary slowdown in the second half based on the investment trend of our major customers, many customers are developing leading-edge devices used in 5G, high-performance computing, IoT, et cetera. This means that there will be wider usage of technologies used for these devices such as miniaturization, stacking and greater integration of circuit on chips, which require more accurate testing and technologies which make reliability more difficult to ensure. We assume these trends should support the growth of the tester market over the medium to long term.
There are uncertainties in the external environment, which is still unresolved such as U.S.-China conflict, the global economy and exchange rate fluctuations. And we cannot be certain about market recovery in Q4 yet. Therefore, we have to be ready for the risk of some changes in market environment. We'll grow Advantest business further as we carefully examine opportunities and risks. We'll continue to strive to achieve the goals set forth in the mid- to long-term management policy grand design by reinforcing our customer support capabilities.
Today, we distributed the integrated report at the entrance. This is the first integrated report, and we appreciate if you go through the report and send your feedback, questions and comment.
Now we would like to receive your questions. Please give your name and affiliation, and please show your hand before you ask your questions.
Your name, please?
Wadaki, Nomura Securities. Thank you. Questions about SoC and memory testers.
First, SoC testers. Market expectations are growing so rapidly. TSMC is receiving a lot of orders in the [ advanced ] area on a short-term basis. So I assume it will have a consequent effect on advanced or on the 5G testers, and they are making big investments. With the current 5G testers going on so strongly, once the mass production starts, SoC and test market may explode. I wonder if you have the same expectations.
So there are the 2 questions. And maybe Mr. Mihashi can respond about TSMC. Then Mr. Sakamoto respond about 5G. And maybe Mr. Yoshida can make a kind of summary.
Thank you for your question. Yes, in the advanced processes, the latest nodes are growing, pushing the growth of tester services. As Mr. Yoshida mentioned this earlier, with the advanced processes growing, naturally, the number of transistors to be incorporated is naturally dramatically growing.
For example, with the latest new 7-nano EUV, it has more than 1 million transistors. And that period was the earlier than 7-nano margin lithography. The number of transistors grew more than 20% or 30%. So with that in the background, naturally, our SoC testers will have a positive impact, together with more transistors and incorporate it, making tests more complex, particularly so now in the latest nodes area. If this trend [ on a -- hits those find raising ] with the latest nodes, SoC demand in the world will naturally grow. So we believe, well, with a possible slowdown for some time to come, we are assuming there could be some adjustments possible in the second half. That's -- and of course, we do believe that demand for our SoC will grow, together with the advancing nodes.
Maybe I did not make myself clear. Let me double check that. And if you also believe that TSMC in 5 nano [ or the 816 ] has not finished yet, it will further come back. What do you think?
Well, in the previous briefing session, we told you that we had received a big order for our node testers. And this has been already covered and announced by the TSMC. I believe you understand what I'm talking about. So the question here is how this mass production affairs ramps up, with the base stations being built in 2020 and onward, 5G-based mobility on the smartphones in the world will naturally ramp up. And they will be in several of our customers, and they are now using TSMC processes. Yes, with this point in mind, as I said this earlier, there will be more demand for the Advantest and SoC in the marketplace.
And you mentioned about 5G. As we explained this in our last financial results briefing, we received orders mainly driven by the engineering. And in the second quarter, we did receive orders from those customers who entered the initial mass production for ES or CS engineering samples or the commercial samples. And as Mihashi just mentioned, those testers prepared are not to be used. So a slight slowdown could be possible in the second half. So there could be somewhat a slowdown. But moving toward 2020, several customers of ours will move into mass production phase in our fourth quarter.
So in 2020, more 5G devices will be made. Together with the infrastructure or the base stations being built, there will be more devices created in the first and the second half in 2020, and the first half in 2021. Furthermore, the device features will be further enhanced year by year. So we believe more advanced test will be required as the year goes on. They have to conduct tests to verify if their devices are fully functioning or not.
For reliability, there will be more testing items, as well as longer testing time required. All in all, we do believe that tester demand will increase surely and steadily.
So both in volume -- both in terms of the stringent test, no comment. I do believe that there is going to be a further increase in this testing space.
Now sub 6 GB is coming up. And of course, the millimeter [ new ] wave will follow. And once millimeter [ new ] wave is here, they have to support a variety of frequencies, resulting in more complexity of devices. That's pushing up more demand for the testing. And we have not changed our view that 5G will definitely drive those demands on a midterm basis.
Mr. Yoshida, do you have anything you would like to add?
Yes. As the adoption of 5G into the society, I do believe that it will take 10 years, so it is yet to come. And we are just standing at the entrance point. Western devices are becoming more complex and the capacity becoming larger, making testing more difficult. And if you believe that, you can assume that they'll be using exactly the same test methods, you're wrong. Definitely, customers are trying to increase test efficiency, and we are here to support their efforts at the same time. So together with complex testing environment, whether or not there's going to be any relationship between their side and our side, probably not. But many devices are going to come into -- deployed in many aspects of society, not only smartphones. But also in factories and also in offices and industrial equipment, and also in healthcare equipment. Everything is going to be connected.
If that's the kind of assumption we're going to make, I think we have to develop a huge amount of devices. And we are here now to try to offer the broadened support in the field of testing in order to support our customers and partners. It's very important for us to receive good feedback from the customers and partners so that we can be up to the point we have to actually deliver to the marketplace.
Second question. Anyone can respond to my second question, please. Memory testers, particularly for China. As you know, they aren't actually making those -- the products, ignoring the patent. So sooner or later, the U.S. will come and get them. There is such a concern, what do you think? And any view as Advantest?
And as for the flash, some manufacturers have stopped production in refracturing. And they have turned a corner, and I'm just wondering if this will have a positive impact on you? Any comments in this regard?
Well, please try to understand that we are not in a position to say anything about the patents. But that said though, our gut feeling is that demand are encouraging to us and has not changed drastically. Well, and, of course, there are risks. Hope you are with me here.
As for tester demand for NAND, in China, our customers have gone through engineering samples, moving on to initial mass production, and then commercial mass production. So customers are making a study in preparation for CapEx. So naturally, there is going to be a demand for this. So we will support this. We are now moving into such a phase.
Actually, talking about NAND, and I was not talking about om China, rather Samsung or Toshiba, who have already turned the corner. I'm wondering if there have been any investments being made there. Yes, DRAM versus NAND. NAND is definitely we're talking ahead of the DRAM. So sooner or later, there will be a demand, so we believe. Any specific inquiries? Well, not yet, if you are talking about the big inquiries. [ You didn't for your detailed response. ]
We'd like to move on to next question, please.
This is Ishino of Tokai Tokyo Research. My first question is that 5G pre-service got started in China. Some are already offering commercially -- commercial services. So I would like to know how you view the demand for 5G devices this year, next year and the year after next. I believe in tester market, the demand will rise in an accelerated fashion. What I have in mind here is huge currencies you had with 4G LTE. I'm talking about your factories in China. Just in 2 years, from 2014, the demand has shot up 2x to 400 million. And I am just wondering if you are ready to meet with such a rapid increase and demand in 5G. So now let me stop here.
Thank you for your question. And I need to clarify now your question, whether or not if I understood your question clearly. If there is a ramp-up in 5G and whether or not Advantest is ready to meet with the demand, whether not if we have any concern in terms of our production capacity? Am I right?
Well, I would like to know your current forecast as to the 5G demand currently, and whether or not you are ready to go for it.
Well, thank you. Well, out of the current smartphone demand being 1.4 billion globally, then our own assumption for 5G is 10%-plus out of 1.4 billion in 2020 coming from smartphone manufacturers. So we do believe that such a migration to 5G is going to take place, particularly in the following year. That's the assumption how we are making.
Then as for our production capacity, with that in mind, yes, you are right. When we are actually working on in 4G LTE, there was some issues. But now as for this 10% growth of 5G, yes, we are ready in cooperation with our partners. We are ready in terms of pipeline flow.
Furthermore, the consensus in Taiwan or in China is 150 million in 2020. And we are also hearing coming from China now, the minimum 200 million is going to be also impossible. And some manufacturers are a bit stronger in telling us that it could be as big as 250 million or up to 300 million. So here, I would like to inquire if Advantest is truly ready to cope with that kind of scenario.
First, whether or not they expect a number of smartphones and our tester business are in a good alignment, well, I have to say that, that is not necessarily the case.
Put in differently, with devices growing 2x, our testers will also grow 2x. Well, may I remind you that same point I talked about earlier. When our customers are ready to move into mass production, they should have improved their testing efficiency. The surely should have improved their yield. So with this in mind, 2x does not necessarily mean 2x on our side.
The fact is that we and the customers will have a good dialogue, and we do receive the growth forecast. Then we also make our preparation. If there is going to be an explosive growth in demand, yes, we are more than happy to respond to such a demand. But as of now, we are not convinced that, that will be the case.
Putting it differently, some of the 5G smartphones are just in name. We are not 100% sure that they are truly qualified smartphones to be declared as true 5G phones. So may I remind you that we're being somewhat careful here.
Next, my second question and the last question. I would like to ask about DDR5 HPC. If I'm not wrong, a new plant will be built in South Korea next year. What is your view on how much this is DDR5 and HPC account for in next year? And again, I need to ask you if you can be up to the rapid increase in demand.
Thank you. Yes, DDR5 and/or LP-DDR5. Yes, it is likely to ramp up in the second half or in the upcoming fourth quarter. The mass demand comes from the higher-end smartphones, talking about DDR. So this has been our strong belief. And here now, we believe there is going to be the [ linear linkage into the new ] phones in question. We believe the volume ramp up now will be quite possible with the 5G migration, how we're talking about earlier.
And in regard to the question concerning our production capacity. As Mr. Yoshida mentioned this point earlier, with the very initial ramp-up of DDR5 launch, efforts need to be made for better quality. And it will come to the tester business with a higher capacity, and we are happy to report to you that, yes, we have enough production capacity in order to meet with those demands.
Yes, please. I am Maekawa from Crédit Suisse. I have 3 questions. My first question is about 5G. An estimated sales of JPY 142 billion for SoC testers in FY 2019, about how much would be 5G-related sales.?
We expected this question, and various figures are running in my head about it. Let me explain the definition of 5G.
5G devices include base stations and some sub 6 GB baseband and APU will be available in the market. We have explained the expected market size for 5G is from USD 200 million to USD 400 million. This is based on the assumption that testing time will be longer due to more complex device functions and more complex protocol verification to cover both 4G and 5G by introducing 5G technologies.
With that, we estimate an increase of USD 200 million to USD 400 million, including other RF such as millimeter wave bands. To answer your question, I should say our estimate of business growth related to more complex 5G devices functions include analysis on how many customers are actually making the tests more complex and longer. Therefore, it is difficult to answer your question by clearly stating how much sales would come from 5G. We have been discussing this issue internally to make a proper answer to this question, and we suppose it would be more appropriate to do so after taking more time to examine this issue carefully.
Okay, then. I will ask differently. In your plan at the beginning of this fiscal year, sales forecast for SoC was JPY 113 billion. And this time, it was revised upward to JPY 142 billion. Can I assume most of the increase of JPY 3 billion will come from 5G?
You can assume a major part of the increase is related to 5G. Let me explain a little more.
As earlier mentioned by Mr. Wadaki, 5G-related semiconductors use the latest processes such as TSMC's 7 nano plus, and the devices using such processes are considered as 5G-related products. With this background, business opportunities for many engineering and initial mass production increased our SoC-related business volume. Did I answer your question?
Okay. Again, about 5G, to increase production of 5G smartphones, it is necessary to increase usage of mid-range devices, and cost is a big issue for mid-range devices. When you talk to AP vendors like Qualcomm and HiSilicon, what is your observation about their cost consciousness for the mid-range AP?
As Mr. Yoshida mentioned earlier, increase of smartphone production does not mean linear growth of testers as a result. What I want to know most is what is the cost target in this regard, and how much test time reduction can be expected for mid-range AP compared to high-end flagship model AP for smartphone.
As some major 5G devices are currently driven by such customers, and fierce competition already exists among customers for mid-range devices since the age of 4G. We have been receiving requests for lower-cost testing, and that is a challenging issue for a long time. Cost pressure is getting higher for testing OSAT used by fabless customers year by year, and this leads to cost reduction request to us. We are always trying to figure out what is the best way to address this issue in a competitive environment. However, we do not want to start price competition. So it is essential to focus on reducing cost of test through shorter test time or supporting capacity utilization rate improvement. As such, we continue to make various proposals in those areas to help customers' cost reduction. We will not change this attitude for 5G products.
Can you tell me your rough estimation on how many more chipsets can be tested in case of mid-range devices?
Let me give you additional explanation. In general, test cost for customers' devices depends on application. But for low-end devices, the test cost is required to be contained around 1%. In case of some applications with high added value, the cost is about 8%. You suggested pressure on the test cost for middle range devices is high, and that is true. But it cannot be explained simply by stating how much is acceptable. For instance, wafer cost is the same regardless high-end or middle-range devices. And the cost is changed by the processes used. Mid-range devices do not use top-notch processes.
So to answer your question, I can say mid-range devices are under stronger cost pressure. And as I said, we are always under pressure to lower the test cost, and we are trying to figure out the way to improve efficiency by discussing with customers to address this issue.
This is my last question, about your market share. In the presentation, Mr. Yoshida said 50% would be maintained, but he sounded rather conservative. Teradyne said the market share is growing both for DRAM and NAND in their recent financial results briefing. What is your view on your market share trend for DRAM, NAND and SoC in 2018, '19 and '20? I understand this totally depends on customers' investment decision, but I'd like to know what is your thought on this matter.
FY '18 was the year we could increase our market share significantly, especially with SoC. We could record remarkable growth to go over 50%. As for memory, we have been in between 50% to 60%. In this area, we are historically not very strong with flash memory, but strong with DRAM. This is because test specification for flash memory can be developed with relatively [ simple procedures ] therefore, profit margin is not very high, and we have not forecast much on this area.
However, recently, more layers are stacked on NAND, and this makes its test more difficult and requires longer time for testing, and test market for NAND is growing as a result. We cannot ignore this trend, and now, we are strengthening this area. Teradyne may say they are growing their share for DRAM. However, we believe we maintain a strong position for DRAM. We assume their claim of increased market share is just because of larger NAND testing market. When demand for DRAM test is back, we are sure our share will grow again. We have been considered weak with NAND, but we are trying to gain ground in this area. We are not satisfied with 50% market share. Instead, we intend to maintain the market share over 50%. Still, it will be very difficult to reach 60% or 70% considering the current competitive landscape. Our competitor is not only one. There are other competitors in Korea, and we may have other competitors in Taiwan and China for low-end devices. Also, some customers develop and make testers by themselves. Under such a circumstance, we cannot be too optimistic and do not expect over 60% market share. As long as we maintain 50% share as minimum, we can secure the #1 position. So our target is to maintain over 50% market share.
As we are running out of time, I'd like to take one last question before concluding the session. The gentleman in the back, please.
I am Yoshikawa from Morgan Stanley. I have 2 questions. First question is about second half outlook. You repeatedly suggested possible slowdown in the second half. Can I assume you just suggested possibility and you do not see any sign of slowdown? Or are you seeing slowing down situation already in October? Regarding smartphone-related business, investment in OSAT could start again around the Chinese New Year. If you are not sure about it, what is your concern?
My second question is aside from smartphone, what is your second half outlook for display driver, automotive devices and MCU? They were weak in the first half, but do you expect recovery in the second half?
For the second half, estimated order and sales are smaller than the first half, and it is natural to have such impression that our business performance will be weaker. However, we still have order backlog for Q3, and we do not expect significant decline in sales. Still, we cannot have a clear outlook for orders in Q3 and Q4. We can see the situation in October. But as November and December are the months when customers lower their production, and with uncertainty of world economy, we do not expect strong order taking during that period.
In Q4, when new fiscal year starts for our customers, especially with memory-related business and in the U.S., it is possible new investment opportunities arise from them. Given that circumstances, our attitude may look too conservative. If market environment changes in Q4, for instance, by an announcement of an agreement between U.S. and China in November, we can expect recovery. We maintain conservative view for now based on the situation we currently observe.
Regarding display driver, automotive devices and MCU, we were not strong in the first half. And for the second half, unfortunately, we still cannot see a sign of strong recovery. Display driver had performed quite well until last year. And taking last fiscal year as 100, this year stands at a level of 50 or 60, and which is not considered as a serious decline. But to be honest, we cannot clearly see when it will come back to 100.
Recently, we heard Korean display driver companies announced significant profit decrease or deficit. And as such, good news cannot be heard very much in this area. Still, in terms of technologies, demand for display driver will grow, and we expect its recovery sometime in next year. As for automotive devices, as concerned all over the world, we cannot expect a sudden recovery in the area. However, electrification is progressing steadily and significant investments are being made in Europe. In that sense, semiconductors are making a recovery. And with increasing importance of reliability, they need to perform thorough testing, which requires testers. We are sure about this medium-term trend and not pessimistic about the future. Unfortunately, though, this trend is not obvious yet.
I recognize some more people are raising hands, but I am sorry, we have run out of time. This concludes the Financial Briefing for the Second Quarter FY 2019 of Advantest Corporation. Thank you once again for participating the briefing despite your busy schedule.