Advantest Corporation
Q3 2021 Earnings Call Transcript

Published: 29 Jan 2021

[Operator:]

Thank you for attending the Earnings Conference Call of Advantest Corporation for Fiscal 2020 Third Quarter Ended December 31, 2020. Participants from Advantest are Yoshiaki Yoshida, President and CEO; Atsushi Fujita, CFO; Kimiya Sakamoto, Executive Vice President of Sales Group; and Yasuo Mihashi, Executive Vice President of Corporate Planning and Stakeholder Relations. First CFO Fujita, will go over the financial results for the third quarter, followed by a presentation on the fiscal 2020 outlook by CEO, Yoshida. We will then take your questions. The entire conference call is scheduled for an hour. Materials are available from TDNet as well as the Company Web site.

[Atsushi Fujita:]

Good afternoon, ladies and gentlemen. This is Fujita speaking, I will go over the results for the third quarter. Please turn to Slide 4. First of all, we would like to express profound gratitude to our suppliers, other business partners and employees, who all continue to support Advantest business amidst a second wave and a third wave of the COVID-19 pandemic. Business results for the October-December period exceeded our prediction made in October. SoCs and semiconductor demand from High Performance Computing or HPC and smartphone sectors was strong, and our business environment improved faster than was expected three months ago. Supported by these trends, the third quarter SOC tester orders as well as overall orders reached new quarterly record. Sales and profits also exceeded expectations. Orders, sales and profit will be explained in detail in the following slides. Next slide please. The third quarter orders by segment, first semiconductor and component test systems, up 49.3% quarter on quarter at JPY66.8 billion. SoC tester orders reached JPY54.2 billion, an increase of JPY18.4 billion from the previous quarter. Display driver ICs or DDI principally drove the orders against the background of a increasing adoption of OLED in 5G smartphones and increasing incorporation of touch sensors into DDI. Demand for application processors for 5G smartphones also grew more than expected, helping to eliminate the surplus of testers in the supply chain affected by the US China trade friction and generating new investment. Orders of the memory testers by Chinese players increased by JPY3.6 billion from the previous quarter. Next Astronics systems, up 29.8% quarter on quarter at JPY11.8 billion. AUV related demand drove higher nanotechnology product orders. Services support and others up 61.8% quarter on quarter at JPY16.5 billion. In addition to a seasonal increase in annual maintenance contract renewals, system level test related orders increased. Next slide. This slide is about Q3 sales by segment. Semiconductor and component tech system sales increased by 9.8% quarter-on-quarter to JPY52.3 billion. Sales breakdowns are JPY39.3 billion for SoC testers, and JPY13 billion for memory testers. Smartphone related products, especially DDI and image sensors, were the driving force behind the increase in SoC tester sales. Sales of memory testers decreased due to a decline in orders in Q2. Megatronics system sales decreased 15.4% quarter-on-quarter to JPY9.7 billion. Sales of test handlers and device interfaces fell in line with that decline in sales of memory testers. Sales of services, support and others declined by 11% quarter-on-quarter to JPY16.3 billion. Q3 sales of system level test products decreased in reaction to stronger momentum in the first half but remained in line with expectations.

[Yoshiaki Yoshida:]

This is Yoshida speaking, I will present the fiscal 2020 outlook. Please turn to Slide 12. First, semiconductor testing market trends. Six months ago, we warned that surplus test capacity generated by the tightening of US regulations on a major Chinese player could significantly reduced demand for testers. However, test demand for HPC has continued to be strong. North American smartphone related test demand has been higher than expected. And competition for market share among Chinese smartphone makers has seasoned the elimination of surplus testers from the supply chain. Respecting these market changes, we have raised our calendar year 2020 SoC tester market estimate to approximately $2.8 billion.

[Operator:]

We will now take questions. First questioner is Mr. Wadaki from Nomura Securities.

[TetsuyaWadaki:]

Thank you for this very strong and encouraging earnings presentation. My first question, the high tech stock prices have plummeted very rapidly. The reason being a comment from a certain semiconductor manufacturing devices manufacturer that it is sensing a pause in inquiries. Do you share that same observation?

[YoshiakiYoshida:]

As far as smartphones are concerned, true, there is an obvious overproduction. So I suspect there will be a pause eventually.

[TetsuyaWadaki:]

So what is your take on this matter?

[YoshiakiYoshida:]

First, it's hard to define what that company means by a pause. As far as a dentist is concerned, as you knew in July, there was a situation which was referred to by some people as Advantest shop, the strong third quarter that we experienced was driven by an upward trend starting in October, December being the highest during the third quarter.

[TetsuyaWadaki:]

Would this trend discontinue in the fourth quarter?

[YoshiakiYoshida:]

Actually, we expect orders to remain strong. So rather than a pause, we feel that strong business is continuing. Having said that, I think it is true that the COVID-19 situation is accelerating the drive for digital transformation. When this acceleration becomes excessive, there naturally will be a correction. But that is not likely over a new term like three months or six months from now but a bit further than that, partly in life as a stimulus packages implemented in different parts of the world.

[TetsuyaWadaki:]

My second question. At the last earnings call, you said that your share vis-a-vis the smartphone manufacturers that are increasing production is smaller than your share within manufacturers that no longer can produce smartphones, such as Huawei or HiSilicon, and that due to such competitive factors, your earnings momentum lags behind your competitors? Now you have made an upward revision to your full year guidance this time around. Is that due to market factor or shift in particular market shares?

[YoshiakiYoshida:]

It's really not a shift in our -- due to the market shares. The way the major US smartphone company operates in terms of investment is that investment activities will be completed more or less by summer. Our market share remains high even with that completion, this is probably because a fairly strong demand continuing elsewhere. So on a customer by customer basis, I think the market share situation remains almost the same. As for supplying into the Chinese market, our market share is fairly high. In terms of smartphone demand in China and other markets, we are maintaining our shares pretty well. I doubt there was a sudden shift in our market share at individual customers.

[TetsuyaWadaki:]

So I take that display driver, ICs and image sensors that were mentioned earlier are supporting your market share, correct?

[YoshiakiYoshida:]

Yes. As far as for display driver, ICs, we have much large market share. So the expansion of the market itself is one sector that makes our orders appear strong. But the way we see it, the overall business is strong, including non-DDI products, such as application processors and CMOS image sensors.

[Operator:]

The next questioner is Mr. Nakamura from Goldman Sachs Japan.

[ShuheiNakamura:]

I have two questions. One is on the outlook of the SoC tester market. You had made an upward revision to the 2020 outlook. As always, can you give us the breakdown by three application areas? I'm particularly interested in what specific areas are expected to grow over the last three months of the fiscal year. You did mention driver ICs and others, but can you touch upon the qualitative aspect as well?

[KimiyaSakamoto:]

As was mentioned earlier, we raised the order forecast by JPY16 billion, of which JPY30 billion is in the third quarter and JPY20 billion in the fourth quarter. In the third quarter, SoC testers accounted for 80% of that upside. So the majority was SoC testers driven by DDIs. And of the JPY20 billion in the fourth quarter, SoC accounted for 50%. Under these circumstances, we increased the forecast of the size of total addressable market TAMP with increase in 5G and smartphone related amounting to $100 million.

[YoshiakiYoshida:]

Mr. Nakamura, I take it that you are interested in the segment breakdown of SoC testers, correct?

[ShuheiNakamura:]

Yes, that is correct.

[YoshiakiYoshida:]

And it's computer communication 70%, automotive, industrial and consumer 30% and display drivers a few percents, that will be the market breakdown.

[ShuheiNakamura:]

So is it fair to say that the market outlook has been revised upwards by $100 million, most of which comes from upward revision for the display driver ICs in accordingly the upward revision in your order forecast for the third and fourth quarters is also attributable to display driver ICs. Am I correct?

[KimiyaSakamoto:]

This is Sakamoto speaking, let me clarify. $100 million revision for TAM is in relation to HPC and AI related. The increase in the third quarter orders that was mentioned earlier was largely driven by display driver ICs.

[ShuheiNakamura:]

My next question is on the outlook for the 2021 for SoC testers and memory testers respectively. You have shared your outlook. Can you describe where you expect growth in terms of application areas? Also, I think from 2020 to 2021, we can expect a recovery in your market share in SoC testers. So can you give us your sense on that.

[KimiyaSakamoto:]

For 2021 the driver will be HPC with continued progress in investment in advanced node processors. We also expect CIS to grow significantly on a calendar year basis. In terms of growth rates, DDI growth is expected to be very large. Since we command a high market share here, we want to make sure we capture that growth in our business. As for memory testers as was mentioned by CEO, Yoshida, we expect a similar level as in 2020 at minimum with a possible increase of around $200 million. So we expect growth primarily driven by DRAM and NAND front end process testers.

[ShuheiNakamura:]

On SoC, earlier you said that at the backdrop of their current order growth is the market share competition amongst smartphone manufacturers. So with that in mind, do you feel that there will be a further growth in 2021, or has part of that expected growth already materialized?

[KimiyaSakamoto:]

Orders levels are very high for the third and fourth quarters. There could be some ups and downs on a quarterly basis, of course. But our conversation with our customers does not indicate or suggest any signs of slow down. So we expect the strong trend to be sustained.

[Operator:]

Thank you. The next questioner is Mr. Hirakawa from BofA Securities Japan.

[MikioHirakawa:]

I have a couple of questions. First is on the gross margin. You said the gross margin in the third quarter stayed low at 51.5% due to unfavorable SoC product mix. What factors give you the confidence in recovery in the fourth quarter? More specifically, can you give us the assumptions for the fourth quarter gross margin? Is it that product mix does not change much between the third and fourth quarters, but as new system level test product that existed in the third quarter will no longer be a factor in the fourth quarter and that is going to make a difference, which is the case? That's my first question.

[AtsushiFujita:]

In the full year guidance, we are expecting the cost of sales assumption in the fourth quarter forecast. As you have correctly described, those with better profitability in terms of cost is to account for larger proportion of sales. And as such, we expect a recovery in the fourth quarter over the third quarter.

[MikioHirakawa:]

I hesitate to ask the specifics, but while there was better profitability is to account for a larger proportion, new system level testing products mentioned as low profitability items. Would that be the biggest factor, or is it that even among SoC testers the proportion of low profitability items is going to go down while high profitability items for application processors are going to go up?

[AtsushiFujita:]

One of the reasons for deterioration in cost of sales rate in the third quarter was that cost the system level test product with some cost accrual taking place in advance in the third quarter. So in the absence of that in the fourth quarter, we expect an improvement. The impact of improvement in system level test is not that significant. Overall, I think you can safely assume that the improvement in product mix is making a difference in gross margin.

[MikioHirakawa:]

My next question. What is your current view system level test market size calendar year 2020 versus calendar year 2021?

[YasuoMihashi:]

We are frequently asked about the SLT market size to which our response always is, it's hard to define and therefore, we can't answer clearly.

[MikioHirakawa:]

I apologize for asking the same question over and over.

[YasuoMihashi:]

Honestly, there's really nothing that we can say at the current moment as to the size of the SLT market.

[MikioHirakawa:]

Let me rephrase the question. In terms of TAM or in terms of the market that you are looking at, do you have any thoughts on what the market size is going forward?

[YasuoMihashi:]

For the market size, we are seeing increasing opportunities to talk with our customers. So the size is growing, the market size is growing, obviously. However, making growth a reality requires iterative conversations with each customer individually. So we are unlikely to see major growth. But our impression is that demand in 2021 will be on par with that of 2020 or slightly higher.

[Operator:]

The next questioner is Mr. Sugiura from Daiwa Securities.

[ToruSugiura:]

I have two questions. First is again on SoC tester market outlook. For 2021, you are showing a range, the lower limit being flat year on year. Previously when we asked for your qualitative comments, I believe your response had a more bullish tone. So this time around, what kind of risk do you have in mind in projecting a market growth?

[YoshiakiYoshida:]

First, let me point out the fact that the business significantly grew from 2019 to 2020, despite we did not expect such growth. Investment by major US smartphone companies contributed this growth and this increased TAM. As for 2021, our assumption is our investment from such interior smartphone companies will slow down but overall market will grow due to increasing demand for other testers. This means there is a chance for us to recover our market share in 2021, which decreased in calendar year 2020 in that sense our SoC testers can be expected to grow higher than the growth of TAM.

[ToruSugiura:]

My second question is about building up your production capabilities explained in the presentation. I suppose you mentioned this on SoC testers and the need to ask a follow up question about it. Can I assume your production system is already getting tight? And as you're going to build up your production system, are there any additional expenses expected for next fiscal year, including CapEx? I appreciate your comments on it.

[YoshiakiYoshida:]

First of all, we have both outsourcing and in house production for SoC testers. And in the third wave of COVID-19, for instance, in Malaysia, a strict regulation was imposed at the beginning of this year. Therefore, we are considering to prepare proper production system as our BCP or business continuity planning, so that we do not have to depend solely on outsourcing partners but this does not require CapEx. Regarding production in Japan, now all are saying that semiconductors are in shortage and in our case, testers need components, such as semiconductors and other electronics product. And it is true that supply of those components is getting more and more tight. Under such a circumstance, orders are increasing significantly and we need finetuned management to fulfill all of these orders and minimize risks of delayed delivery to our customers. Such potential risks are contained by our daily efforts. However, if COVID-19 situation gets worsened or any bottlenecks are generated in the global supply chain, we assume there is a risk this affects other areas. In that sense, we will secure system to supply properly. As I said before, our idea is not to expand out all the factories for production, rather we'll manage outsourcing partners properly and we do not consider additional capital investment here. CapEx investment has increased slightly for this year but this does not mean to build new factories. This is to be used for other facilities to develop new customer or to own testers in our facilities. Therefore, we do not estimate much CapEx for production increase.

[Operator:]

Next, Mr. Yoshida from CLSA.

[YuYoshida:]

My first question is about your outlook on orders. Based on my calculation with the revised full year forecast, it seems Q4 orders will decrease Q3. Could you tell us what direction you expect about orders by product type? Also, according to the earlier explanation, Q3 performance got better towards the end and the good performance still continues. And if that is the case I assume you take some potential risks into consideration in this declining Q4, I want to know about it as well.

[YoshiakiYoshida:]

Q4 order is estimated to be nearly JPY90 billion, and this is less than Q3 but we consider this is still at higher level. The quarter-on-quarter difference is due to the decrease in DDI related orders, following the extremely high demand for DDI in Q3.

[YuYoshida:]

Do you mean demand for DDI is expected to decrease slightly in Q4?

[YoshiakiYoshida:]

Yes, the decrease is in reaction to the high demand for DDI in Q3.

[YuYoshida:]

And other areas are expected to have flat growth basically, correct?

[YoshiakiYoshida:]

Yes, we estimate flat growth for computer communications, automotive and industrial machinery.

[YuYoshida:]

I have a follow-up question about automotive and industrial machinery. It seems automotive does not have much orders at this moment but are there any new inquiries? Earlier you did not mention much about recovery in automotive when you talk about SoC tester market outlook. Can you tell me your outlook in this area?

[YoshiakiYoshida:]

A lot of news are going on about automotive business and tester utilization rates among our automotive customers have risen sharply and we are receiving new inquiries. Having said that, in quarter on quarter comparison, its growth is still limited compared to HPC and smartphone related businesses. So for 2021, customers are increasing inquiries in consideration of production growth and we have a confidence for the year.

[YuYoshida:]

In that sense do you mean it is possible that orders for automotive application becomes stronger, although, it was not mentioned in the earlier explanation?

[YoshiakiYoshida:]

Yes, 2021 will be surely stronger than 2020.

[YuYoshida:]

Here's my second question. I hear an MPU manufacturer is considering to outsource production to foundries. I think this company was once your customer but switched to in house made customized testers. Do you think it is possible to get back your business with the company as a result of their outsourcing to foundries?

[YoshiakiYoshida:]

We refrain from commenting on an individual company. So we cannot make any comments. To be honest, we do not have enough information to comment at this moment.

[YuYoshida:]

Do you think if you can have business with the company, it could benefit HPC business more than your current estimate?

[YoshiakiYoshida:]

We have no idea. We don't know anything about the business flow, for instance, where they are going to perform testing. So we are not in position to comment on it.

[Operator:]

Thank you. Next, Mr. Hasegawa from Mitsubishi UFJ Morgan Stanley.

[YoshihitoHasegawa:]

First, let me clarify your production system. In your integrated annual report, it is said the X is the key. Can you elaborate more on this? And considering a substantial order backlog at present is your production system is working properly enough to address such backlog? I appreciate if you explain about it. This is my first question.

[YoshiakiYoshida:]

Regarding our production system, we said we forecast over JPY80 billion in sales for Q4 and if you quadruple the number, it will be JPY320 billion. We believe our current production system is good enough to support such sales under ordinary circumstances. However, we have concerns explained today related to COVID-19 and we are not certain about how much impact is expected by the virus, such as lockdown or component procurement failures as a result of production suspension at factories due to a sudden surge of infection in a certain country. But we think we are okay if there are no such major supply chain disruptions. But in reality, we faced the situation that the factory in Malaysia, which supplies SoC testers for us, suspended its operations for about a week at the beginning of this year. Therefore, we are building the new production system in a German subsidiary so that it can take over the operations in other countries if required. This is our BCP. Also, the production system in Japan is well established to adjust any situations for products made in Japan as long as we have enough component supply. Therefore, we are sure that we can address product sales over JPY80 billion even at present.

[YoshihitoHasegawa:]

Can I assume that capacity as well as production efficiency are improving substantially?

[YoshiakiYoshida:]

Production efficiency of which location?

[YoshihitoHasegawa:]

Mainly for domestic operations.

[YoshiakiYoshida:]

Are you asking if it is worsening due to COVID-19?

[YoshihitoHasegawa:]

No. I mean, in terms of advances of manufacturing, you mentioned various strategies in the integrated annual report. So my question is, if you are improving not only operations but also productivity at your domestic factories in terms of progress on the initiatives?

[YoshiakiYoshida:]

Production costs and sales costs have not changed much. So, it is difficult to say we have improved the efficiency. I think you know if you have visited our factory before that we have not introduced many advanced machines, but we started initiatives last year to improve production efficiency by using Robert and various data. Although, it is not achieved in the current mid term management plan, we expect the factory production system will be much advanced in the next midterm management plan. So my answer is, our production efficiency is not worsened even with higher production volume and we are making efforts to improve the efficiency.

[YoshihitoHasegawa:]

Let me ask my second question. I'm sorry to ask the same question every time, but what is the progress in data analysis solution business? I appreciate any updates on it as long as you can disclose.

[YasuoMihashi:]

You mean the business you are collaborating with PDF. Based on the development with PDF Solutions announced earlier, we’ll prepare cloud environment, which is the foundations to extend data in the future and put internal tools on it to stop service deployment. With the current devices, such as the ones with 5 nanometer process node, data to be taken from semiconductor production process for analysis by customers have already exceeded one petabyte, that is a data volume expected. And some obstacles still exist in various silos from semiconductor processing to device testing. We are trying to connect all of these silos to improve productivity by effective use of all data and we are still in the preparation to build this framework.

[YoshihitoHasegawa:]

When the partnership was announced, I think I heard you already had started working for a customer. Are you facing more obstacles than you originally thought in this progress?

[YasuoMihashi:]

We are making progress and activities with the specific customer but we do not set a goal of this initiative in those activities. We are still a company mainly focused on hardware business. So we are trying to figure out what we can do as we discussed with various customers to know how we should drive data into production efficiency or productivity, which are issues for customers and add value to improve time to cash.

[Operator:]

This is the end of the conference call. Thank you very much for your participation, and please make sure you disconnect the call.