May 28, 2019 Medium-term Management Plan Briefing Q&A

For each of the three scenarios in the new financial model on page 12 of the presentation materials, what size of WFE*1 market are you presuming there will be?

Although the WFE market is currently in an adjustment phase, our viewpoint from last year that it will exceed $60 billion in the mid- to long-term has not changed. However, if the WFE market is linked to the financial model, then the discussion will be centered on market share rather than profitability. Furthermore, with the belief that we will be able to create a new SAM by integrating our technology while new demands arise in the future such as HPC demand, this year we announced a model that places operating margins at the forefront rather than the WFE market share.
As to our financial model’s timeframe, the model announced last year uses the next fiscal year, FY2021, as the target. This year, we have judged that rather than announcing the FY2021 model like last year, it is better to show what kind of profitability we will aim for within the next five years.
Another change we made to our financial model was setting the target ROE to at least 30%. We set a range of 30%-35% for the target ROE last year, but, this time, we removed that upper limit. The message we would like to make with this change is that, while taking into account necessary growth investment and how cash should be used depending on the situation, we will achieve a ROE of at least 30%.

When was the SAM*2 outlook prepared? Was it prepared after taking into account the current uncertainty in matters such as US-China trade friction? And what is your viewpoint on WFE, which is what SAM is premised on?

It was created approximately four weeks ago and represents mostly up-to-date figures. We, of course, referenced WFE trends. However, SAM forecast for the next year would be more accurate if the forecast were made based on information we have compiled on the scale of customers' investments. Elements such as market adjustments and US-China trade friction are difficult to forecast, so we are not taking them into account. However, investments by local Chinese manufacturers are not yet large, so, for the next two-to-three-year period, we see the influence of US-China trade friction as somewhat limited.

When the financial model was announced last year, the targets for the gross profit margin, selling, general & administrative expenses, number of days for accounts receivable turnover, and number of days for inventory turnover were disclosed. Could you tell us the reason why you didn’t announce those this time?

The target figures for this year are not greatly different from those announced last year.

I would like to ask about the penetration rate of comprehensive service agreements in the Field Solutions business. What is the current situation and the target figure for FY2023? Furthermore, for results-based compensation services, what will happen if TEL is not able to achieve a KPI set with a customer?

Around half of the customers that we are aiming to execute a package service agreement have executed some kind of package services agreement. However, the percentage of customers who have executed a comprehensive services agreement in which high value is added, such as assured uptime through using TELeMetrics™ or AI, is less than 10%.
As to results-based compensation services requirements, they differ depending on the agreement executed with the customer. If we do not meet a KPI, then we incur a burden, such as, for example, extension of the warranty period.

Page 70 of the presentation materials show that the etch systems SAM for CY2020 is forecast to grow by around 30% year-on-year. Has there already been some customer activity suggesting an increase by 30%? Or will there be some activity in the second half of the year?

This is difficult to forecast, but currently we are forecasting that in around CY2019Q3 (i.e., July to September), customers will perhaps start to be active.

With respect to the sales forecast for the etch system on page 72 of the presentation materials, is the large swing up in HARC processes*3 sales due to the memory investment?

For CY2019, as the memory market has a bearish tone, the sales for the HARC processes will decrease.

The presentation materials state that an etch system SAM share of 30%–35% will be the target. Could you tell us what the primary factor is behind this range? I would like to hear what the upside and downside risks are. Also, isn’t the lower limit of 30% the same level as the current level?

It certainly looks as though that 30% means that we will just stay as we are today, but we are not overestimating the downside risks. In the current year, CY2019, our outlook is that both the market size and HARC processes sales will decrease, so the business environment will be severe and we will maintain a level of 30%. With a long-term view, however, we would like to increase the level to close to 35% through growth in the HARC processes and the patterning processes.

Could you tell us what the current state of progress is with respect to the new etch system for memory?

We are developing a system that incorporates new functions for the HARC processes. For the 3D NAND 128-layer generation and beyond, we would like to capture a share of channel processes with this new system.

What is the reason for your belief that you can capture a POR*4 for HARC processes? Is it because your process performance is excellent? Or is it because the process time is short? Also, what is your current share in the market for HARC processes?

We are aiming to capture a POR with a proposal that satisfies both process performance and productivity requirements. We don’t disclose the share for individual processes.

In the presentation materials, it is forecast that the SAM for the etch system’s HARC processes will grow enormously. Is this growth forecast based on the assumption that the necessary number of chambers will increase due to the longer process time? For the 3D NAND 128-layer generation, how much longer is the process time compared to the 64-layer generation?

The number of layers will double. However, with deeper holes, the etch productivity will drop, so the process time will be longer at over double the time.

For the 3D NAND for the 25X generation and beyond, are you calculating the SAM under the presumption that most customers will adopt two stacks?

We calculated the SAM forecast under the presumption that all customers will adopt two stacks from the 25X generation onwards.

The announcement on the share buyback was the first since April 2015. What triggered you to decide to carry out this buyback? Also, could you tell us your approach with respect to the future cash position and share buybacks?

It was the result of taking into account factors such as growth investment, working capital and our confidence in future growth. Furthermore, we decided to carry out a share buyback as we believe that our potential is not sufficiently reflected in the current share price.
As to our future cash position, it will change depending on the circumstances—for example, in the phase in which we can expect growth, the necessary procurement value might increase. Our policy of considering share buybacks in a flexible manner is unchanged.

What assumptions did you use for cash generation, necessary cash on hand, and growth investment capital for the next three years to determine the share buyback upper limit of 150 billion yen? In particular, I’d like to know what your approach is to growth investment capital.

we would perhaps like to allocate around 1/3 of adjusted operating income to each of the following three categories: (1) development/growth investment, (2) return to shareholders and (3) return to employees and working capital.

In the past, you commented that you would like to strive to carry out share buybacks of a scale and at a time that would exert an impact on the stock market. Is that policy still valid? Or has your message changed such that, rather than attaching importance to the share price in the short term, you are aiming to reach a ROE of 30% by curbing the increase of shareholders’ equity by continually carrying out share buybacks?

As I’ve mentioned in the past, when considering share buybacks, the share price is also a fairy large factor. The semiconductor industry has somewhat of a cycle, so, rather than carrying out a share buyback at the same value each year, we believe it is better to carry it out at an appropriate time, share-price level and scale.

WFE (Wafer fab equipment): The semiconductor production process is divided into front-end production, in which circuits are formed on wafers and inspected, and back-end production, in which wafers are cut into chips, assembled and inspected again. Wafer fab equipment refers to the production equipment used in front-end production and in wafer-level packaging production

SAM: Served available market

HARC (High aspect ratio contact) process: A process for forming holes that requires advanced processing technology

POR (Process of record): Certification of the adoption of equipment in customers' semiconductor production processes

The above content is a summary of question and answers session. An audio recording synched to the slides is available here.